Some copy-protection schemes exploit asymmetries in the system. For example, there was a system for protecting analogue audio by having an additional frequency that would overload tape-to-tape recording systems but would not interfere with direct playback.
With digital media it is much harder, because you don't need to process the information in order to read it. With analogue, the two steps were the same. Thus, copy protection at the data level is completely useless. If you can read the data, you can indeed copy it.
Copy protection is NOT, however, totally impossible. You just have to approach the problem differently. Instead of protecting at the logical layer, you need to protect at the physical layer. One way to do that is to have disks that can be written to and read on conventional equiptment, but exploit some additional property that is physically added to hold more data than could be stored otherwise, where that property could not be trivially added by pirates.
For example, have a physical masking plate over the disk, such that when the laser burns data onto the disk, only half of the normal area is marked. You then lay down another set of data, marking the remainder of the disk. Finally, you etch the surface in such a way that the laser is focussed onto the half you're wanting it to read at that time.
Your disk can be read on a normal system - the optics are altered only on the disk itself, not in the player - but there's now twice as much data on the disk than can be written without that masking plate. So long as the data is compressed enough that another 50% loss would be unacceptable, the medium becomes readable but effectively uncopyable.
"Effectively" is important, here, because of course the medium is perfectly copyable. You just use the same method as used to create the disk in the first place. The problem would be purely one of obtaining masks capable of giving you the effective density, then obtaining etching technology capable of giving you double-density reads. This is so far beyond Joe Average that piracy would simply not happen.
The "normal" things that concern "real" people - backing up, reading on unsupported OS', etc, would be completely unimpaired by this method. You're not altering people's ability to read the data, you're only altering their ability to duplicate the disk as a disk.
Why isn't something like this being done? Because piracy isn't the issue. It never has been. As 98% of desktop machines run Windows, 98% of piracy on the desktop is going to be under Windows. Yet interviews with the MPAA have always implied that it was DVD readers on Linux and other (relatively speaking) minority OS' that were their concern. Bit-copiers and frame-grabbers have never been at issue and (through pricing) have been over-compensated for anyway.
The best thing that the F/OSS world can do is to exploit the time gap between now and when Blu-Ray and HD-DVD come out, to do something better. There are not enough blue-frequency players out there to make a difference, right now, so the market can be effectively regarded as open. Since there are no movies long enough to take advantage of the extra capacity (other than LoTR), such disks will only be of interest for data storage anyway, for now.
It should not be hard to produce a read/write optical disk with greater capacity than either of these schemes propose. Flat lenses, polarised light, diffraction gratings, three colours of laser (red, green and blue) - there's got to be something trivial enough to tag on that would nonetheless make the system unique enough not to fall foul of patents AND sufficiently better to make Blu-Ray and HD-DVD too primitive to fall back to, when they do come out.
The best way to pwn the market is to define the market. You're never going to convince the RIAA and MPAA, so don't involve them in the first place. If they're runners-up, then they'll have to make do with what YOU choose to provide.
Whilst dragons (especially the Pernese kind) are probably a little beyond the engineering of the day, I see no reason why you couldn't have something that was a good substitute. The Haast Eagle had a wingspan of 14' and could certainly have carried a small person. As far as anyone can tell, it's about the upper limit of purely muscle-powered flight of that kind, so it could not be used to get adults to work. On the other hand, it might be a way to get kids to school.
(It was also a bird of prey, so detention for more troublesome kids would not be necessary.)
We already have talking turkeys, but I won't say which reality TV show they're on. Seriously, although you'd never be able to get a bird to talk, the avian brain is demonstrably capable of a respectable level of intelligence. African Grey parrots can understand grammar, adjectives and context, whilst crows have been filmed manufacturing tools (the only animal other than humans to demonstrate manufacture, not just exploitation of pre-existing resources).
The problem with PETA has nothing to do with their arguments or beliefs. Their problems revolve around communication skills and a lack of hard data - both of which could be fixed. Slogans do not an argument make (did Yoda say). Do animals have intelligence? Do animals have feelings? We have fMRI, we could settle this beyond dispute. By making ethics an issue of faith alone, those with different faiths can - and do - feel free to ignore those ethics. Once ethics becomes a matter of verifiable fact, faith is irrelevent and can say what it likes.
The communication aspect is just as important. Fine, let's say some things are wrong, but society (for whatever reason) feels they're important. Animal experimentation being a good (or bad) example. Can you communicate why the existing option really isn't as good as it seems -and- offer a better, cheaper, easier alternative? The alternatives do exist, but if you can't communicate that convincingly, why should anyone care? Society has inertia and unless you're willing to be convincing enough, inertia will always win out.
Phew! Ancient manuscripts! Thanks for the reference. On the one hand, we obviously have a lot more empirical data now, some of which may well conflict with earlier theories. On the other hand, the mechanics of organisms - for the most part - was certainly well within the science of the day and it's unlikely that there will be any significant discrepencies except for the more extreme structures.
(It is unlikely that the dynamics of twenty tonne rhinos, or fourteen foot eagles, would have been within range of consideration. In 1917, flight was still too new for them to have been able to extrapolate accurately, and the scalability of power systems would have been uncertain. Everything else should hold up just fine.)
Yeah, things have changed in computing. My PET 3032 had, well, 32K of RAM. So did the BBC micro. Fast-forwarding a bit, it is still possible to run Linux in 4 megs of RAM (2, according to some), so it would be possible to load a working image - kernel and userspace - into the L2 cache of this chip. No need for main memory at all, then. It would suck on performance, but without the need of any support chips or main memory, you'd have one hell of an embedded system.
Actually, one of my favourite processors from the late 80s - the transputer - had onboard memory for exactly that purpose. You could slap on 4 gigs of external RAM if you wanted, but you could actually throw quite respectable applications into the CPU itself and never have to worry about data busses again.
Other posters have given excellent comments on the inner workings aspect. I'll throw in a few more, though, just for luck.
First, the fastest overclocked Intel processor was something like 7 GHz, so it is evident that the electronics are capable of substantially better performance.
Second, the new Intel chips are hyperthreaded and multicore......but don't obviously allow the different cores to access the full set of processing elements, only the ones in that core. It would seem easy enough to have the actual processing element pool split from the main part of the core, so that all processing elements are available to all cores.
The biggest limitation is moving data around, not the CPU itself. Adding HyperTransport, DMA, etc, to the CPU itself would be a Good Thing, as would doubling the width of the data bus.
Executing all possible paths is not efficient when combined with hyperthreading, as you're wasting processing elements. Probabalistic branching (where you only follow one branch, but it's the most likely) would seem more efficient and would free up more elements for better threading.
I don't believe registers are ping-pong buffered, but it would save having to wait on writes if you still need to do reads because of the difference in timing from multithread execution.
Nobody uploads microcode to CPUs, but everybody runs code that would be efficient if run internally on the CPU. It would be good if the OS could upload atomic architecture-specific hardware operations into the CPU as pseudo instructions. Save having to hunt through physical memory for common tasks that will likely fall out of cache if you rely on that.
Processor overheating is a big problem and keeps the speed down. Processor casing simply isn't optimal for keeping the internals cool. It wouldn't be hard to improve the heat transfer from the chip surface to the casing surface.
Processors aren't made from optimal materials. If you're using silicon, for example, you want something that is single isotope, stressed and allowed to crystalize slowly. It's substantially cheaper to produce flawed silicon wafers, but they will never perform as well.
Along with this, I've learned that the reason aluminium is the most popular for CPU interconnects and copper is second, with silver unused, is problems with silver being too reactive and copper being only just managable. If they could find a way to prevent the silver from reacting with the rest of the CPU - should be possible - then you'd improve speeds there, too.
Electron leakage is a problem, as it also imposes a speed limit. Not sure how you'd prevent it, but there might be ways to limit the problem. Electrons have spin. It is certainly possible to polarize something by spin, and it is certainly possible to filter by polarization. There MAY, therefore, be ways to limit the impact of leakage and therefore ways to bypass the speed limit such problems would otherwise cause.
The only difference, in this case, is that rodents actually scale very well (as the above two examples show) and are designed in very similar ways, so the scaling on one (the two listed are closest to gerbils and hamsters) should work reasonably well on another (such as rabbits).
In general, though, you are absolutely correct about scaling. An insect doesn't scale well - partly because of limbs, but mostly because they don't have a scalable way to absorb oxygen. The 60s sci-fi horror of giant mutant insects is a physical impossibility. The proportions are all wrong and they'd die from a lack of oxygen.
Likewise, elephants seem to be fairly limited in size. There are no fossil records of a pint-sized elephant - although there ARE fossil records of a pint-sized horse.
Rhinos scale well - the largest in that family reaching 18' in height. It is not clear how small they can get, but I seem to recall that the Manatee is a fairly close relative that returned to the oceans.
Humans seem to be about the upper limit on size - the largest of the Great Apes was only about 10' - but fossil finds seem to indicate that hominids have existed which were much smaller. I'd call hominids passably scalable, then.
I'd love to see someone do some in-depth research on what forms ARE scalable and where this differs from what has actually occured in practice, why that potential might not be viable in practice.
The largest rodent that lived was the Phoberomys pattersoni, weighing in at 1500 lbs and was the size of a buffalo. Amblyrhiza Inundata was a good runner-up, the size of a black bear.
All you need are the genes controlling size from these, and you could have a single rabbit for the entire block! Much more efficient and you could possibly train it to keep guard at nights.
...this will be achieved by extracting DNA from the bones of the Dire Wolf, the Bone-Crushing Dog and the Epicyon then genetically embedding the fragments into a poodle. Aside from the fact that it will then have three ears and meow on thursdays, it will be much placated with the therapy.
Polarized light video displays (3D in color, and it's pretty damn good, even if you do need polarized glasses) in the home
DVD vendors finally concede defeat and make their products genuinely interchangeable, even when using a home recorder
SCO completes the transformation into a Ringwraith and adds Frodo to the lawsuit
Someone develops a completely functional computer that runs Linux and Fedora Core, using only chip specifications from Open Cores and programmable components. The computer then outsells at least one well-known PC manufacturer.
SGI reaches crisis point and can't continue. It is bought by OSDL, the Altix is moved to the Opteron and Linus Torvalds sets a new record for kernel build times.
...that's because something in his test suite breaks with SP2. Be interesting to know!:)
Seriously, there are many freebie compression tools which weren't mentioned but which are in common enough use that they can be regarded as highly significant in the market, or which are simply SO good that they are likely to become significant. Zzip and SZip are big ones that didn't get mentioned.
Further, since speed is considered, it is unfair to list bzip2 without mentioning pbzip2 or bzip2smp (two parallel versions), as you'd obviously get a speed boost from non-sequential compression. Not sure what it does to the compression ratio, though.
Finally, some forms of compression - notably Huffmann Compression - rely on the size of the compression table to determine how well you'll compress data. On a modern computer, where multiple gigs of RAM is no longer unusual, you could reasonably look at frequencies for 24-bit strings.
Most Huffmann compression will use 8-bit frequency tables, a few will use 16-bit, because the memory requirements get big. Fast. Not only do you need to record the total frequency of the wordsize you're using, you also need to have space to build the encoding tree. Even then, the level of compression you'll get will only improve to a point. After that, longer words will produce worse compression or even inflation.
In the examples used - audio and video data - you will most likely want a 16-bit word for the audio and a 24-bit word for the video, because that reflects the nature of the data. 8-bit words on the frequency tables are going to be crap, because you're compressing random fragments of words, so artificially worsening the encoding tree you're going to build.
I have two points here. First is that by picking the right (or wrong!) parameters for the data, you can always rig a benchmark. My second point is that you can often tailor an algorithm that would normally be worse than some other algorithm such that the worse of the two will outperform the default behavior of the better one.
Ideally, you'd use a form of arithmetic encoding, but that is so riddled with patents that although you could (in theory) develop a system which was numerically identical but did not infringe on the wording of the patents, most Open Source and low-cost vendors don't bother trying.
The secret, in compression, is not to use default algorithms if you can avoid it. (Ideally, this would be when the compression header stores enough information for you to tweak table sizes, etc, so that off-the-shelf decoders will work with your custom encoders.)
The Presidency is largely picked, these days, by whoever can grab the most money and most airtime. Neither of these qualities would seem to be particularly meritorious in deciding who should run the country. The President can rule for 4 to 8 years, with questionable oversight. The first term is spent solely concerned with being re-elected. The second term is spent solely in becoming a historical figure. Neither term is spent in doing anything useful.
Congress is also decided by whoever has the most money, together with where the political boundaries are that week. (See: DeLay and, indeed, most other boundary cherry-pickers.) Again, not exactly redeeming qualities.
The judges are currently political picks by the President (see above) and approved by Congress (see above). As demonstrated, time and again, nominees are rarely honest during the interview process, making any approval process a farce anyway. Because there is no transparency, justice cannot be seen to be done. In high profile cases (eg: O. J. Simpson), the judge seems more interested in the book deals afterwards rather than in maintaining any semblance of sanity in the courtroom.
In all of these cases, I cannot see how a random pick could be any worse than the existing system. It's much more likely to be proportionally representative of the population, so arguably could present a better case of representing them. By reducing terms to the absolute minimum that can still be productive and useful, you eliminate the existing situation where stagnation and inertia are killing all rational debate, and where lobbyists can control virtually all aspects of Government.
Maybe you'd need the President to stay in office for more than a few months, to keep international affairs in any kind of order, but the current system has become a really bad joke. The jury pool system is bad, but it's actually the most functional, most respected, most honest segment of the entire governing system that we have.
The laws were enforced meaningfully, evenly and without discrimination
The punishments educated minors and adults as to why there might be a problem, rather than being disciplinarian, arbitrary and meaningless
The warnings that did exist (such as for smoking) were not paranoid scare-tactics (which are hopelessly ineffective), factually incorrect (which destroys any message there might be) or irrelevant to those concerned (and thus ignored by the very people involved)
Every time I see an ad campaign that is clearly and obviously (to anyone with half a brain cell or more) going to have the exact opposite effect to the one intended, I seriously wonder how mankind managed to get so far yet remain so woefully and obnoxiously stupid.
I do believe that laws that restrict smoking, porn, violent games, etc, can be made to work, work well, and work in a way that can near-universally be agreed upon as good, sensible and mature, even by the most anti-legislative, pure-blooded libertarians out there. I also believe such laws won't come from backroom deals, religious viewpoints and righteous rage. They are far more likely to come from rational and open discussions.
This law on violent games, for example, was brokered by politicians FOR politicians. The judge noted that no correlation between violence and games had been proven. Why could Californian legislators not wheel out neurologists with fMRI studies that could prove a unquestionable cause-and-effect on the mechanical level? Why could they not produce child psychologists who could produce solid, verifiable, repeatable evidence of a correlation on the behavioral level? If they'd done that, what objection could have been raised to there being some response?
They didn't, for an obvious reason. They never talked to any. They never had any data to work from, so had no data to present.
Ok, assuming we now have data that a response is required, we would now have to determine what kind of response is needed. The only people who can tell you what computers can do would be computer experts. The only people who can tell you what businesses can do would be business experts. For parents, you probably want to talk to a mix of parents and sociologists.
They didn't do any of that, either.
Once you've all that information to hand, you can distill it into a law that has a clear, firm, rational foundation that has unquestionable merit in dealing with a provable and proven problem, in a manner most likely to produce a verifiable and socially beneficial response.
Ah, well, rational legislation seems to be way beyond what we have come to expect from government. A pity, as they have no excuse whatsoever in producing anything else.
This assumes legislation is needed at all, of course. If the neurologists cannot show a mechanism AND the child psychologists cannot show that said mechanism produces an actual, verifiable response that is adverse and mentally toxic, and which cannot be avoided by changing some other parameter, then there's nothing for a law to do.
(You have to have both. Just showing a mechanism isn't enough, if the mechanism can be trivially ignored by most people. Even the response is not enough, if you cannot prove beyond all reasonable doubt what triggers it, OR if there were some other change - better education, for example - that could do the job better and more universally.)
The ONLY valid legislation would be IF the science justified legislation in the first place AND the legislation honored what the science defined as being the REAL problem, AND the legislation honored what the experts said society could reasonably respond to, AND the legislation honored what the Constitution defined as being the place of legislation, no matter what the data might say.
If all those conditions had been respected and met, I seriously doubt anybody would have had
The total lack of any power (or right) to do anything, all it would be able to do is observe. The McCarthy trials were tainted because you mixed investigative powers with judicial powers with political powers with media access. That's WAY too much power, concentrated in one place.
There were other past abuses along those lines. The Watergate scandal, for example, mixed espionage with political power with domestic law enforcement.
My argument is that you should rip ALL investigative powers from the executive and legislative branches. Diffuse the power. Break the ability for one group or another to combine powers, because that is almost invariably a recipe for disaster.
If investigators could NOT be controlled politically at all, they could not be used as a political weapon. If their terms were measured in weeks or (at most) a few months, the poisoning atmosphere would not have long enough to take hold. Furthermore, as their terms would be so short, there would be no suckering-up and no loyalty to personality cults. Indeed, it would be hard for a personality cult to form in such a short time. Thus, investigators would be open to investigation themselves and there would be no Good Old Boys Club to protect their backs.
I freely admit that it is entirely possible my idea isn't worth beans, but even in the worst possible case, I just can't see how it could be used in a dictatorial fashion if it couldn't DO anything. The biggest crimes of the NSA (that anyone knows about) are some industrial espionage in Germany, the clipper chip saga, the current political spying fiasco and other political spying fiascos. In other words, they all involve either outside control or control of the outside.
If you eliminate control from the equation (in either direction) and require absolute single-purpose, single-function operation, you would render improbable (if not impossible) any of these scenarios. Control is where intelligence has always gone wrong and it is in intelligence that control has always gone wrong. Divorce the two. Totally. They have no business being linked, as the Iraq WMD scandal has shown all too clearly.
Once those who control cannot control the information, they can either co-exist or be a slave to it. They have no other choices. If those who possess the information cannot control it either, they too must co-exist or be a slave to it. As nobody chooses to be a slave - well, other than some really strange people, but there are other boards for such discussions - the only option left is a more mutual coexistence with reality.
As one might gather from that last paragraph, I think a reduction in secrecy has to be involved. Ownership of information is dangerous, particularly if linked to a "need to know", as "need" is never going to be defined from a functional standpoint, only from a manipulative one. You'll never eliminate secrecy, but the more transparent the system and the more dispersed the data, the less powerful the data will be.
Information is like the rings of power, in the Lord of the Rings. If concentrated too much, like the One Ring, it can be ruinous. If dominated, like the Nine, it can be fatal. Only when independent and distributed, like the Three, can it be used meaningfully. Even then, it was a razor's edge deal in the books and not much better in real life. Hence the extra need for transparency. The analogy isn't perfect, and I'm bad at analogies anyway, but I hope that conveys the importance of getting information away from those with the power to use it, as much as is humanly possible.
The idea is very close to the notion of Blitzkreig (what the US calls "Shock and Awe"), but would have no teeth. Quite deliberately. It would be an extremely dangerous idea to have something like that, if it could actually DO something. By totally disarming it, it would essentially be a lightning-reaction force of investigative journalists with guaranteed inside tracks on the scandals at the price of ONLY being allowed to anonymously self-publish for zero profit and working a temp-agency style contract.
Power corrupts, but it takes time. Therefore, if you have only the briefest instant of time, you can only have the smallest degree of corruption. Also, if they can't DO anything, then most of the advantages power provides that allow corruption to occur are severely crippled.
I think it could be made to work, in a way that really would be a good idea. I also think it is inevitable that something analogous to this approach will evolve, sooner or later, because Governments are the least able to police themselves but are the most in need of being policed. The only questions in my mind are: (1) how, in fact, this will come about, and (2) when. If you look at every system that has failed in history, it is because corruption has spread from the center, precisely because Government is powerless to monitor itself. Ergo, if you want a failure-resistant system, you have to find a way to add monitoring.
Would be that it is picked from the Federal jury pool, with each person remaining in office between 3-6 months maximum, with all of the usual anonymity that usually goes along with juries. It's sufficiently long for a reasonable investigation, but sufficiently short that neither threats nor promises would hold much weight.
Actually, I'll make a small alteration to that. Those doing the collecting (ie: not actually processing anything, they're just doing the grunt work) could be kept on for 5 days maximum. That would be sufficient to get badges prepared, brief the raiders, get them to the site, raid the place, get them back, gather the materials found and debrief them. After that, their anonymity is gone anyway, so they'd be much less effective and much more open to bribes.
Keeping things running flat-out and then cycling things as rapidly as possible should keep both the politicians and corporations out of it. They'd not have enough time to react, before there had been sufficient turnover to make whatever they were reacting to moot.
Most of the screens had been built using an old version of X-Designer and replicated code excessively. The original code had maybe a few hundred near-identical screens, each made from a few dozen near-identical panels, each and every single panel for each and every screen being individually coded. The code also individually coded menus, event handling, etc.
What I did was pick out the generics of what was being done. That was hard-coded into a function. The specifics of what was needed for any given instance was passed in as a parameter. Each screen was reduced to a handful of function calls and some data.
How does this apply to Government? Well, I'd say start with the basic principle of what I did. You've a library of generic functions (which is basically what top-level departments are supposed to be) and data for the specifics (which is what the subdivisions are supposed to provide). But because this is treating them as a library of functional components, you have to invert the power structure. Atomic operations should always be treated as LOW level.
Functions common between departments should be pulled out and made common. That should be the job of the civil service. It's what they are there for. So, eliminate ALL duplication and centralize that in your political C library.
Ok, what about my suggestion for a new branch? How does that fit in? Well, it serves the same purpose as having an error handler. Trapping error conditions that cannot be trivially caught in advance is a Good Thing, but a program cannot do this inline efficiently or effectively. There will be too many cases where such errors will occur outside of an inline handler. It is also massively redundant, as most such checks will be identical and duplicated across most of the code.
Error checking in Government is handled inline and is massively duplicated. A very substantial portion of Government is tied up in accounting, for example. For reasons of secrecy, many of the departments that are the most wasteful also have the poorest accounts and billions simply vanish. The DoD and intelligence departments routinely ignore the GAO and Senate. They know they'll get the money they want, regardless. The DoE isn't much better, especially as much of energy policy is decided in strictest secrecy. As for the House and Senate ethics committees - internal investigations only ever turn up what is politically expedient. You wouldn't get far if criminal trial juries could only ever have criminals on them.
No, it would be better to have one group with absolute, unquestionable access to EVERYTHING, regardless of security considerations, that handled ALL the accountability (financial and legal). You eliminate those functions ENTIRELY from each and every department and replace the IRS with a department whose sole purpose is general-purpose Government I/O - such as delivering tax information and collecting money. You can also eliminate the GAO entirely.
Redundancy is the biggest contributor to bloat. Unnecessary functions come a close second. We've so far eliminated a lot of redundancy, by inverting the power structure (so primitives don't have more power than the high-level operations driving them) and by moving a common, generic operation that is too prone to be corrupted when inline into an independent thread.
Once you've done that, Government should appear lighter, as you don't have all the heavyweight stuff at the top and you've properly threaded the exception handlers. Now, are there any unnecessary functions we can get rid of as well? Well, if the Senate adopted Proportional Representation or some similar system, then you could eliminate the role of Vice President. Why have a person whose sole real job is to tie-break, when you can just change the system to one that can't tie?
In fact, at this point, it is unclear what role a President would have. Most countries have a Prime Minister, who is merely the head of one of the houses, so would be the Majori
I've found the best way to fix bloatware in a program is to add support for debugging, monitoring and valgrinding. Once you've done that, then ripping out the dead parts of the code is easy. Until you've done that, then you can't be sure if what you're removing is important or not, or if it simply needs writing better.
One program I had to de-bloat was about 15 million lines long, most of it very badly maintained Motif GUI code. I added a 1,000 line widget set to the code, and was then able to remove 14 million lines of unnecessary redundancy.
Adding the right stuff, therefore, CAN lead to removal of the wrong stuff. Adding a pin to a balloon will remove bloat.
As another poster noted, free speech is only one civil liberty. In fact, it's such a minor one, it doesn't even get a mention in the basic framework of the Constitution -OR- the Declaration of Independence. It is only in the first amendment (ie: it's an additional civil liberty) that we get to hear about freedom of speech.
Your core civil liberties - the ones your ancestors fought for in the Revolutionary Wars - were "life, liberty and the pursuit of happiness". Nothing about speech there. And all three CAN be taken away by terrorists, and HAVE been taken away by the current administration*.
*The administration has carried out policies of assassination, incarceration without charge or review, and intimidation of selected political and ethnic groups. That would seem to violate all three of the core civil liberties on which the Constitution is built. These core civil liberties are not protected within the Constitution, presumably for the same reason you don't put the foundations inside the building. But if you erode those, the whole structure is going to fall down. Remove the foundations and you remove everything.
Unfortunately, Dubya also understands what it means to have majorities in both houses, soon to be the Supreme Court as well. It means that the odds of an impeachment crossing BOTH houses AND surviving an appeal would be next to nil. Especially as he is popular with the extreme right and it's the extreme right that'll probably decide 2006' elections. After that point, impeachment proceedings would last longer than the remaining presidential term.
In short, there's absolutely nothing anybody can do about him. There are no effective safeguards and no meaningful counterbalances for this kind of situation. The best any moderate can hope for is that both the 2006 and 2008 elections are decided by great enough margins towards those who want effective safeguards, that it'll be as easy to stabilize and secure the system then as it has been for the current administration to corrupt it.
My personal preference would be for a constitutional amendment that added a wholly new branch of Government - outside the Executive, Legislative and Judicial - that has all the necessary powers, clearances, means and protections to investigate corruption at absolutely any level in every branch of Government. That is it. That is all it would do. Just investigate. Because it was independent of all other branches, it would not have political appointments made to it, could not be ordered to stop, or indeed even ordered to start. The power of such a body is not in what it could do, but in what it could know.
Government is often corruptible, not because it is powerful - most humans are powerful over something in their lives, but aren't necessarily abusive - but because few in Government have any reason to believe anyone'll know about it. The moment you can guarantee that (a) someone WILL know about it - no matter how classified the information, and (b) they're utterly protected against reprisals if they talk, then those in power will be much less likely to step over the line. (And, if they do genuinely feel as though they have to, they're going to put every ounce of effort into establishing WHY no alternatives are viable, because they WILL be asked questions later.)
Minimum hardware will depend on the quality and the level of security that you want. I'd say that a mid-range Pentium IV should be sufficient for fairly respectable results on both, and even a 386DX could produce unencrypted telephone-grade VoIP, so a 486DX should be able to encrypt the stream at a very basic level.
Software depends on how you're intending to do the encryption. If you're planning on hooking up to a regular phone system as well as doing VoIP, then you're certainly looking at using Asterisk for your exchange system.
At the kernel level, for pure VoIP, you probably want to use either Linux with either the StrongSWAN or OpenSWAN patches applied, OpenBSD or MirBSD. (I believe FreeBSD and NetBSD have IPSec, but I'm not sure.) This allows you to encrypt from your machine to the destination, using a grade of encryption that will be proof against standard wiretaps. Rijndael (AES) is (as far as anyone knows) uncrackable on existant technology and if you combine it with SHA-2, you've a system that would be impervious to any wiretap you're likely to encounter.
For working with Asterisk, you want to use a stream cipher, not a block cipher. Asterisk with Encryption WIKI has more information on how to set it all up. You basically link up a VPN and encrypt the VPN. Because some line loss is inevitable, you want an algorithm that is resistant to loss and is reasonably fast. There are phones that have hardware encryption in them, which provide higher levels of security and they are also listed.
Of the stream ciphers out there, FEAL/SEAL and Chameleon seem to be the most reputable.
A direct ancestor of mine - James Duncan - restored the Garden of Pamplemousses in 1849. There's very little historical record of the man, or of what he did there, but they do have a nice plaque mentioning him. Somewhere. Unless something has eaten it.
Mauritius is also famous for having a great many highly endangered species. (Think numbers in the single digits.) They're also infamous of having released some, after rebuilding the population in captivity, only for the locals to devour them back out of existence.
The entire island is heavily overpopulated with humans, the environmental concerns are badly neglected (or deliberately overlooked), yet the biodiversity that still exists - albeit in captivity in hopelessly underfinanced, understaffed, underequipped shelters* - exceeds that of virtually the entire USA combined.
*By this, I do not mean they can't go and buy a billion dollar gadget once a week. Accounts usually depict the staff involved in saving the native species as being borderline starvation cases, constantly under threat of one kind or another - particularly of malicious closure, and probably earn less in a month than a kid in a Nike sweatshop can make in an hour.
It is precisely because of conditions like that that I am not the least bit surprised that Dutch scientists would have smuggled out anything they could. You know what? I'm not the least bit sorry for Mauritius. When you treat scientists with enough contempt - hey, scientists are human too, and even the most professional will eventually return the compliment.
I doubt anything will change - except maybe for the worse - in my lifetime. My only hope is that there is something left to salvage by the time attitudes change.
Just invent "anti-lights" - christmas lights 180' out of phase with everyone else's christmas lights. The light will cancel out, and the homes will be saved. You'll make a fortune selling them, not to mention going on talk shows. Your only concern will be the rabid scientists with uzis coming after you for breaking the laws of physics. (You're not supposed to be able to set up interference patterns with essentially random non-coherent light sources.)
Slashdot Mirror
With digital media it is much harder, because you don't need to process the information in order to read it. With analogue, the two steps were the same. Thus, copy protection at the data level is completely useless. If you can read the data, you can indeed copy it.
Copy protection is NOT, however, totally impossible. You just have to approach the problem differently. Instead of protecting at the logical layer, you need to protect at the physical layer. One way to do that is to have disks that can be written to and read on conventional equiptment, but exploit some additional property that is physically added to hold more data than could be stored otherwise, where that property could not be trivially added by pirates.
For example, have a physical masking plate over the disk, such that when the laser burns data onto the disk, only half of the normal area is marked. You then lay down another set of data, marking the remainder of the disk. Finally, you etch the surface in such a way that the laser is focussed onto the half you're wanting it to read at that time.
Your disk can be read on a normal system - the optics are altered only on the disk itself, not in the player - but there's now twice as much data on the disk than can be written without that masking plate. So long as the data is compressed enough that another 50% loss would be unacceptable, the medium becomes readable but effectively uncopyable.
"Effectively" is important, here, because of course the medium is perfectly copyable. You just use the same method as used to create the disk in the first place. The problem would be purely one of obtaining masks capable of giving you the effective density, then obtaining etching technology capable of giving you double-density reads. This is so far beyond Joe Average that piracy would simply not happen.
The "normal" things that concern "real" people - backing up, reading on unsupported OS', etc, would be completely unimpaired by this method. You're not altering people's ability to read the data, you're only altering their ability to duplicate the disk as a disk.
Why isn't something like this being done? Because piracy isn't the issue. It never has been. As 98% of desktop machines run Windows, 98% of piracy on the desktop is going to be under Windows. Yet interviews with the MPAA have always implied that it was DVD readers on Linux and other (relatively speaking) minority OS' that were their concern. Bit-copiers and frame-grabbers have never been at issue and (through pricing) have been over-compensated for anyway.
The best thing that the F/OSS world can do is to exploit the time gap between now and when Blu-Ray and HD-DVD come out, to do something better. There are not enough blue-frequency players out there to make a difference, right now, so the market can be effectively regarded as open. Since there are no movies long enough to take advantage of the extra capacity (other than LoTR), such disks will only be of interest for data storage anyway, for now.
It should not be hard to produce a read/write optical disk with greater capacity than either of these schemes propose. Flat lenses, polarised light, diffraction gratings, three colours of laser (red, green and blue) - there's got to be something trivial enough to tag on that would nonetheless make the system unique enough not to fall foul of patents AND sufficiently better to make Blu-Ray and HD-DVD too primitive to fall back to, when they do come out.
The best way to pwn the market is to define the market. You're never going to convince the RIAA and MPAA, so don't involve them in the first place. If they're runners-up, then they'll have to make do with what YOU choose to provide.
(It was also a bird of prey, so detention for more troublesome kids would not be necessary.)
We already have talking turkeys, but I won't say which reality TV show they're on. Seriously, although you'd never be able to get a bird to talk, the avian brain is demonstrably capable of a respectable level of intelligence. African Grey parrots can understand grammar, adjectives and context, whilst crows have been filmed manufacturing tools (the only animal other than humans to demonstrate manufacture, not just exploitation of pre-existing resources).
The problem with PETA has nothing to do with their arguments or beliefs. Their problems revolve around communication skills and a lack of hard data - both of which could be fixed. Slogans do not an argument make (did Yoda say). Do animals have intelligence? Do animals have feelings? We have fMRI, we could settle this beyond dispute. By making ethics an issue of faith alone, those with different faiths can - and do - feel free to ignore those ethics. Once ethics becomes a matter of verifiable fact, faith is irrelevent and can say what it likes.
The communication aspect is just as important. Fine, let's say some things are wrong, but society (for whatever reason) feels they're important. Animal experimentation being a good (or bad) example. Can you communicate why the existing option really isn't as good as it seems -and- offer a better, cheaper, easier alternative? The alternatives do exist, but if you can't communicate that convincingly, why should anyone care? Society has inertia and unless you're willing to be convincing enough, inertia will always win out.
(It is unlikely that the dynamics of twenty tonne rhinos, or fourteen foot eagles, would have been within range of consideration. In 1917, flight was still too new for them to have been able to extrapolate accurately, and the scalability of power systems would have been uncertain. Everything else should hold up just fine.)
Actually, one of my favourite processors from the late 80s - the transputer - had onboard memory for exactly that purpose. You could slap on 4 gigs of external RAM if you wanted, but you could actually throw quite respectable applications into the CPU itself and never have to worry about data busses again.
...the French are planning on using the new Intel chips to trigger fusion reactions purely from the heat output...
First, the fastest overclocked Intel processor was something like 7 GHz, so it is evident that the electronics are capable of substantially better performance.
Second, the new Intel chips are hyperthreaded and multicore...
The biggest limitation is moving data around, not the CPU itself. Adding HyperTransport, DMA, etc, to the CPU itself would be a Good Thing, as would doubling the width of the data bus.
Executing all possible paths is not efficient when combined with hyperthreading, as you're wasting processing elements. Probabalistic branching (where you only follow one branch, but it's the most likely) would seem more efficient and would free up more elements for better threading.
I don't believe registers are ping-pong buffered, but it would save having to wait on writes if you still need to do reads because of the difference in timing from multithread execution.
Nobody uploads microcode to CPUs, but everybody runs code that would be efficient if run internally on the CPU. It would be good if the OS could upload atomic architecture-specific hardware operations into the CPU as pseudo instructions. Save having to hunt through physical memory for common tasks that will likely fall out of cache if you rely on that.
Processor overheating is a big problem and keeps the speed down. Processor casing simply isn't optimal for keeping the internals cool. It wouldn't be hard to improve the heat transfer from the chip surface to the casing surface.
Processors aren't made from optimal materials. If you're using silicon, for example, you want something that is single isotope, stressed and allowed to crystalize slowly. It's substantially cheaper to produce flawed silicon wafers, but they will never perform as well.
Along with this, I've learned that the reason aluminium is the most popular for CPU interconnects and copper is second, with silver unused, is problems with silver being too reactive and copper being only just managable. If they could find a way to prevent the silver from reacting with the rest of the CPU - should be possible - then you'd improve speeds there, too.
Electron leakage is a problem, as it also imposes a speed limit. Not sure how you'd prevent it, but there might be ways to limit the problem. Electrons have spin. It is certainly possible to polarize something by spin, and it is certainly possible to filter by polarization. There MAY, therefore, be ways to limit the impact of leakage and therefore ways to bypass the speed limit such problems would otherwise cause.
In general, though, you are absolutely correct about scaling. An insect doesn't scale well - partly because of limbs, but mostly because they don't have a scalable way to absorb oxygen. The 60s sci-fi horror of giant mutant insects is a physical impossibility. The proportions are all wrong and they'd die from a lack of oxygen.
Likewise, elephants seem to be fairly limited in size. There are no fossil records of a pint-sized elephant - although there ARE fossil records of a pint-sized horse.
Rhinos scale well - the largest in that family reaching 18' in height. It is not clear how small they can get, but I seem to recall that the Manatee is a fairly close relative that returned to the oceans.
Humans seem to be about the upper limit on size - the largest of the Great Apes was only about 10' - but fossil finds seem to indicate that hominids have existed which were much smaller. I'd call hominids passably scalable, then.
I'd love to see someone do some in-depth research on what forms ARE scalable and where this differs from what has actually occured in practice, why that potential might not be viable in practice.
All you need are the genes controlling size from these, and you could have a single rabbit for the entire block! Much more efficient and you could possibly train it to keep guard at nights.
...this will be achieved by extracting DNA from the bones of the Dire Wolf, the Bone-Crushing Dog and the Epicyon then genetically embedding the fragments into a poodle. Aside from the fact that it will then have three ears and meow on thursdays, it will be much placated with the therapy.
Seriously, there are many freebie compression tools which weren't mentioned but which are in common enough use that they can be regarded as highly significant in the market, or which are simply SO good that they are likely to become significant. Zzip and SZip are big ones that didn't get mentioned.
Further, since speed is considered, it is unfair to list bzip2 without mentioning pbzip2 or bzip2smp (two parallel versions), as you'd obviously get a speed boost from non-sequential compression. Not sure what it does to the compression ratio, though.
Finally, some forms of compression - notably Huffmann Compression - rely on the size of the compression table to determine how well you'll compress data. On a modern computer, where multiple gigs of RAM is no longer unusual, you could reasonably look at frequencies for 24-bit strings.
Most Huffmann compression will use 8-bit frequency tables, a few will use 16-bit, because the memory requirements get big. Fast. Not only do you need to record the total frequency of the wordsize you're using, you also need to have space to build the encoding tree. Even then, the level of compression you'll get will only improve to a point. After that, longer words will produce worse compression or even inflation.
In the examples used - audio and video data - you will most likely want a 16-bit word for the audio and a 24-bit word for the video, because that reflects the nature of the data. 8-bit words on the frequency tables are going to be crap, because you're compressing random fragments of words, so artificially worsening the encoding tree you're going to build.
I have two points here. First is that by picking the right (or wrong!) parameters for the data, you can always rig a benchmark. My second point is that you can often tailor an algorithm that would normally be worse than some other algorithm such that the worse of the two will outperform the default behavior of the better one.
Ideally, you'd use a form of arithmetic encoding, but that is so riddled with patents that although you could (in theory) develop a system which was numerically identical but did not infringe on the wording of the patents, most Open Source and low-cost vendors don't bother trying.
The secret, in compression, is not to use default algorithms if you can avoid it. (Ideally, this would be when the compression header stores enough information for you to tweak table sizes, etc, so that off-the-shelf decoders will work with your custom encoders.)
Congress is also decided by whoever has the most money, together with where the political boundaries are that week. (See: DeLay and, indeed, most other boundary cherry-pickers.) Again, not exactly redeeming qualities.
The judges are currently political picks by the President (see above) and approved by Congress (see above). As demonstrated, time and again, nominees are rarely honest during the interview process, making any approval process a farce anyway. Because there is no transparency, justice cannot be seen to be done. In high profile cases (eg: O. J. Simpson), the judge seems more interested in the book deals afterwards rather than in maintaining any semblance of sanity in the courtroom.
In all of these cases, I cannot see how a random pick could be any worse than the existing system. It's much more likely to be proportionally representative of the population, so arguably could present a better case of representing them. By reducing terms to the absolute minimum that can still be productive and useful, you eliminate the existing situation where stagnation and inertia are killing all rational debate, and where lobbyists can control virtually all aspects of Government.
Maybe you'd need the President to stay in office for more than a few months, to keep international affairs in any kind of order, but the current system has become a really bad joke. The jury pool system is bad, but it's actually the most functional, most respected, most honest segment of the entire governing system that we have.
Every time I see an ad campaign that is clearly and obviously (to anyone with half a brain cell or more) going to have the exact opposite effect to the one intended, I seriously wonder how mankind managed to get so far yet remain so woefully and obnoxiously stupid.
I do believe that laws that restrict smoking, porn, violent games, etc, can be made to work, work well, and work in a way that can near-universally be agreed upon as good, sensible and mature, even by the most anti-legislative, pure-blooded libertarians out there. I also believe such laws won't come from backroom deals, religious viewpoints and righteous rage. They are far more likely to come from rational and open discussions.
This law on violent games, for example, was brokered by politicians FOR politicians. The judge noted that no correlation between violence and games had been proven. Why could Californian legislators not wheel out neurologists with fMRI studies that could prove a unquestionable cause-and-effect on the mechanical level? Why could they not produce child psychologists who could produce solid, verifiable, repeatable evidence of a correlation on the behavioral level? If they'd done that, what objection could have been raised to there being some response?
They didn't, for an obvious reason. They never talked to any. They never had any data to work from, so had no data to present.
Ok, assuming we now have data that a response is required, we would now have to determine what kind of response is needed. The only people who can tell you what computers can do would be computer experts. The only people who can tell you what businesses can do would be business experts. For parents, you probably want to talk to a mix of parents and sociologists.
They didn't do any of that, either.
Once you've all that information to hand, you can distill it into a law that has a clear, firm, rational foundation that has unquestionable merit in dealing with a provable and proven problem, in a manner most likely to produce a verifiable and socially beneficial response.
Ah, well, rational legislation seems to be way beyond what we have come to expect from government. A pity, as they have no excuse whatsoever in producing anything else.
This assumes legislation is needed at all, of course. If the neurologists cannot show a mechanism AND the child psychologists cannot show that said mechanism produces an actual, verifiable response that is adverse and mentally toxic, and which cannot be avoided by changing some other parameter, then there's nothing for a law to do.
(You have to have both. Just showing a mechanism isn't enough, if the mechanism can be trivially ignored by most people. Even the response is not enough, if you cannot prove beyond all reasonable doubt what triggers it, OR if there were some other change - better education, for example - that could do the job better and more universally.)
The ONLY valid legislation would be IF the science justified legislation in the first place AND the legislation honored what the science defined as being the REAL problem, AND the legislation honored what the experts said society could reasonably respond to, AND the legislation honored what the Constitution defined as being the place of legislation, no matter what the data might say.
If all those conditions had been respected and met, I seriously doubt anybody would have had
There were other past abuses along those lines. The Watergate scandal, for example, mixed espionage with political power with domestic law enforcement.
My argument is that you should rip ALL investigative powers from the executive and legislative branches. Diffuse the power. Break the ability for one group or another to combine powers, because that is almost invariably a recipe for disaster.
If investigators could NOT be controlled politically at all, they could not be used as a political weapon. If their terms were measured in weeks or (at most) a few months, the poisoning atmosphere would not have long enough to take hold. Furthermore, as their terms would be so short, there would be no suckering-up and no loyalty to personality cults. Indeed, it would be hard for a personality cult to form in such a short time. Thus, investigators would be open to investigation themselves and there would be no Good Old Boys Club to protect their backs.
I freely admit that it is entirely possible my idea isn't worth beans, but even in the worst possible case, I just can't see how it could be used in a dictatorial fashion if it couldn't DO anything. The biggest crimes of the NSA (that anyone knows about) are some industrial espionage in Germany, the clipper chip saga, the current political spying fiasco and other political spying fiascos. In other words, they all involve either outside control or control of the outside.
If you eliminate control from the equation (in either direction) and require absolute single-purpose, single-function operation, you would render improbable (if not impossible) any of these scenarios. Control is where intelligence has always gone wrong and it is in intelligence that control has always gone wrong. Divorce the two. Totally. They have no business being linked, as the Iraq WMD scandal has shown all too clearly.
Once those who control cannot control the information, they can either co-exist or be a slave to it. They have no other choices. If those who possess the information cannot control it either, they too must co-exist or be a slave to it. As nobody chooses to be a slave - well, other than some really strange people, but there are other boards for such discussions - the only option left is a more mutual coexistence with reality.
As one might gather from that last paragraph, I think a reduction in secrecy has to be involved. Ownership of information is dangerous, particularly if linked to a "need to know", as "need" is never going to be defined from a functional standpoint, only from a manipulative one. You'll never eliminate secrecy, but the more transparent the system and the more dispersed the data, the less powerful the data will be.
Information is like the rings of power, in the Lord of the Rings. If concentrated too much, like the One Ring, it can be ruinous. If dominated, like the Nine, it can be fatal. Only when independent and distributed, like the Three, can it be used meaningfully. Even then, it was a razor's edge deal in the books and not much better in real life. Hence the extra need for transparency. The analogy isn't perfect, and I'm bad at analogies anyway, but I hope that conveys the importance of getting information away from those with the power to use it, as much as is humanly possible.
Power corrupts, but it takes time. Therefore, if you have only the briefest instant of time, you can only have the smallest degree of corruption. Also, if they can't DO anything, then most of the advantages power provides that allow corruption to occur are severely crippled.
I think it could be made to work, in a way that really would be a good idea. I also think it is inevitable that something analogous to this approach will evolve, sooner or later, because Governments are the least able to police themselves but are the most in need of being policed. The only questions in my mind are: (1) how, in fact, this will come about, and (2) when. If you look at every system that has failed in history, it is because corruption has spread from the center, precisely because Government is powerless to monitor itself. Ergo, if you want a failure-resistant system, you have to find a way to add monitoring.
Actually, I'll make a small alteration to that. Those doing the collecting (ie: not actually processing anything, they're just doing the grunt work) could be kept on for 5 days maximum. That would be sufficient to get badges prepared, brief the raiders, get them to the site, raid the place, get them back, gather the materials found and debrief them. After that, their anonymity is gone anyway, so they'd be much less effective and much more open to bribes.
Keeping things running flat-out and then cycling things as rapidly as possible should keep both the politicians and corporations out of it. They'd not have enough time to react, before there had been sufficient turnover to make whatever they were reacting to moot.
What I did was pick out the generics of what was being done. That was hard-coded into a function. The specifics of what was needed for any given instance was passed in as a parameter. Each screen was reduced to a handful of function calls and some data.
How does this apply to Government? Well, I'd say start with the basic principle of what I did. You've a library of generic functions (which is basically what top-level departments are supposed to be) and data for the specifics (which is what the subdivisions are supposed to provide). But because this is treating them as a library of functional components, you have to invert the power structure. Atomic operations should always be treated as LOW level.
Functions common between departments should be pulled out and made common. That should be the job of the civil service. It's what they are there for. So, eliminate ALL duplication and centralize that in your political C library.
Ok, what about my suggestion for a new branch? How does that fit in? Well, it serves the same purpose as having an error handler. Trapping error conditions that cannot be trivially caught in advance is a Good Thing, but a program cannot do this inline efficiently or effectively. There will be too many cases where such errors will occur outside of an inline handler. It is also massively redundant, as most such checks will be identical and duplicated across most of the code.
Error checking in Government is handled inline and is massively duplicated. A very substantial portion of Government is tied up in accounting, for example. For reasons of secrecy, many of the departments that are the most wasteful also have the poorest accounts and billions simply vanish. The DoD and intelligence departments routinely ignore the GAO and Senate. They know they'll get the money they want, regardless. The DoE isn't much better, especially as much of energy policy is decided in strictest secrecy. As for the House and Senate ethics committees - internal investigations only ever turn up what is politically expedient. You wouldn't get far if criminal trial juries could only ever have criminals on them.
No, it would be better to have one group with absolute, unquestionable access to EVERYTHING, regardless of security considerations, that handled ALL the accountability (financial and legal). You eliminate those functions ENTIRELY from each and every department and replace the IRS with a department whose sole purpose is general-purpose Government I/O - such as delivering tax information and collecting money. You can also eliminate the GAO entirely.
Redundancy is the biggest contributor to bloat. Unnecessary functions come a close second. We've so far eliminated a lot of redundancy, by inverting the power structure (so primitives don't have more power than the high-level operations driving them) and by moving a common, generic operation that is too prone to be corrupted when inline into an independent thread.
Once you've done that, Government should appear lighter, as you don't have all the heavyweight stuff at the top and you've properly threaded the exception handlers. Now, are there any unnecessary functions we can get rid of as well? Well, if the Senate adopted Proportional Representation or some similar system, then you could eliminate the role of Vice President. Why have a person whose sole real job is to tie-break, when you can just change the system to one that can't tie?
In fact, at this point, it is unclear what role a President would have. Most countries have a Prime Minister, who is merely the head of one of the houses, so would be the Majori
One program I had to de-bloat was about 15 million lines long, most of it very badly maintained Motif GUI code. I added a 1,000 line widget set to the code, and was then able to remove 14 million lines of unnecessary redundancy.
Adding the right stuff, therefore, CAN lead to removal of the wrong stuff. Adding a pin to a balloon will remove bloat.
...your house is insured for more than it's really worth. The airline company would probably pay you to keep quiet, anyway.
Your core civil liberties - the ones your ancestors fought for in the Revolutionary Wars - were "life, liberty and the pursuit of happiness". Nothing about speech there. And all three CAN be taken away by terrorists, and HAVE been taken away by the current administration*.
*The administration has carried out policies of assassination, incarceration without charge or review, and intimidation of selected political and ethnic groups. That would seem to violate all three of the core civil liberties on which the Constitution is built. These core civil liberties are not protected within the Constitution, presumably for the same reason you don't put the foundations inside the building. But if you erode those, the whole structure is going to fall down. Remove the foundations and you remove everything.
In short, there's absolutely nothing anybody can do about him. There are no effective safeguards and no meaningful counterbalances for this kind of situation. The best any moderate can hope for is that both the 2006 and 2008 elections are decided by great enough margins towards those who want effective safeguards, that it'll be as easy to stabilize and secure the system then as it has been for the current administration to corrupt it.
My personal preference would be for a constitutional amendment that added a wholly new branch of Government - outside the Executive, Legislative and Judicial - that has all the necessary powers, clearances, means and protections to investigate corruption at absolutely any level in every branch of Government. That is it. That is all it would do. Just investigate. Because it was independent of all other branches, it would not have political appointments made to it, could not be ordered to stop, or indeed even ordered to start. The power of such a body is not in what it could do, but in what it could know.
Government is often corruptible, not because it is powerful - most humans are powerful over something in their lives, but aren't necessarily abusive - but because few in Government have any reason to believe anyone'll know about it. The moment you can guarantee that (a) someone WILL know about it - no matter how classified the information, and (b) they're utterly protected against reprisals if they talk, then those in power will be much less likely to step over the line. (And, if they do genuinely feel as though they have to, they're going to put every ounce of effort into establishing WHY no alternatives are viable, because they WILL be asked questions later.)
Software depends on how you're intending to do the encryption. If you're planning on hooking up to a regular phone system as well as doing VoIP, then you're certainly looking at using Asterisk for your exchange system.
At the kernel level, for pure VoIP, you probably want to use either Linux with either the StrongSWAN or OpenSWAN patches applied, OpenBSD or MirBSD. (I believe FreeBSD and NetBSD have IPSec, but I'm not sure.) This allows you to encrypt from your machine to the destination, using a grade of encryption that will be proof against standard wiretaps. Rijndael (AES) is (as far as anyone knows) uncrackable on existant technology and if you combine it with SHA-2, you've a system that would be impervious to any wiretap you're likely to encounter.
For working with Asterisk, you want to use a stream cipher, not a block cipher. Asterisk with Encryption WIKI has more information on how to set it all up. You basically link up a VPN and encrypt the VPN. Because some line loss is inevitable, you want an algorithm that is resistant to loss and is reasonably fast. There are phones that have hardware encryption in them, which provide higher levels of security and they are also listed.
Of the stream ciphers out there, FEAL/SEAL and Chameleon seem to be the most reputable.
Mauritius is also famous for having a great many highly endangered species. (Think numbers in the single digits.) They're also infamous of having released some, after rebuilding the population in captivity, only for the locals to devour them back out of existence.
The entire island is heavily overpopulated with humans, the environmental concerns are badly neglected (or deliberately overlooked), yet the biodiversity that still exists - albeit in captivity in hopelessly underfinanced, understaffed, underequipped shelters* - exceeds that of virtually the entire USA combined.
*By this, I do not mean they can't go and buy a billion dollar gadget once a week. Accounts usually depict the staff involved in saving the native species as being borderline starvation cases, constantly under threat of one kind or another - particularly of malicious closure, and probably earn less in a month than a kid in a Nike sweatshop can make in an hour.
It is precisely because of conditions like that that I am not the least bit surprised that Dutch scientists would have smuggled out anything they could. You know what? I'm not the least bit sorry for Mauritius. When you treat scientists with enough contempt - hey, scientists are human too, and even the most professional will eventually return the compliment.
I doubt anything will change - except maybe for the worse - in my lifetime. My only hope is that there is something left to salvage by the time attitudes change.
The dodo had an important role to play.
Just invent "anti-lights" - christmas lights 180' out of phase with everyone else's christmas lights. The light will cancel out, and the homes will be saved. You'll make a fortune selling them, not to mention going on talk shows. Your only concern will be the rabid scientists with uzis coming after you for breaking the laws of physics. (You're not supposed to be able to set up interference patterns with essentially random non-coherent light sources.)