Jade is a semi-precious lump of rock. Whilst UIDs greater than 4 digits are lumps, the relationship to rock (one of the few good music genres) or preciousness (they don't even LOOK like a Ring of Power) is uncertain at this time.
The poster didn't want SQL, so I'm guessing they ruled out SQLite on those grounds. QDBM and Oracle/Sleepycat/Berkley DB-4 would be the smallest, fastest "pure" flat-file databases that are also cross-platform - at least, that I know of.
With the evolution of very low power, high resolution MRI systems, that's entirely possible. Highly sensitive sensors are generally cheaper than 2.5T (and would be a LOT cheaper than 9.2T) magnets and accompanying shielding. On the other hand, any new field is going to take time to move from early uses to mainstream, and getting sensors to the limits of existing MRI technology is going to take a whole lot longer. (It's one thing to demonstrate basic imaging - which is quite good enough for many purposes. It's quite another to get to the point of imaging down to the level of individual neurons. I'm guessing the 12T magnets used in animal experiments provide far more detail yet.)
...there may be some categories of "inoperable" brain tumours that are inoperable because humans have too low a level of precision. Such tumours would be removable by such a method. There have been many advances in tele-surgery since early work in the early 1990s (Surgeons in Russia operated on patients in America, for example) but this is definitely a lot further forward than might have been expected from the pioneering efforts.
But you CAN test a design, and you CAN validate that a given block of code is functionally identical to a formal design for that same block, ergo for some cases you CAN use testing methods to (indirectly) validate locking even when a direct test would not be possible.
is, in fact, just part of a secret government conspiricy. The REAL constitution is in the archives of Torchwood, in Cardiff, and reads: "Nyah! What's up doc?"
I'm more interested in knowing why, when they plotted the path of the chainmail it looked like suit of, well, chainmail. This is a clear indication of the fractal nature of reality, such that our terminology for spam actually infects the upper layers of reality to the point where even couch potatos are zombified into following this ancient topology.
You have dealt with the logical side of the brain well enough, but what about the emotional side? Children and teenagers especially allow their logical minds to be overridden by their emotions, although adults do the same often enough. (The other reply so far to your post mentions religious nuts. These are people who not only allow emotions to control thought, but demand that emotions do so all the time in everyone. The same is true of the majority of "dedicated" followers of specific politicians or other public figures, or who get passionate about their beliefs on talk radio.) This is why intellectual maturity is not enough. It must be there, but that alone is not sufficient.
The emotional side of the brain is of critical importance, especially because that is the side of the brain that reacts strongest to pleasure, to hormones and to the senses. Oh, I'm sure there are people - especially amongst the geeks - who can get aroused by a paper on superstring theory or who reach the dizzying heights of ecstacy from filling in a gap in the proof of Poincaire's Conjecture. However, that is probably not very common.
Now, geeks are not known for building up emotional maturity very quickly, and pre-science mindsets don't build up intellectual maturity very quickly. If you required both to exceed the age of consent, the population will plummet. Personally, I couldn't give a damn if the pre-science lot died off, but I'm going to be fair then I guess any two out of the three maturity indexes could be considered sufficient. However, if the only mindset we can easily examine is the intellectual one, then that leaves a lot of combinations that should be legal according to common sense left illegal. If we want to create (or just imagine) a wholly rational, equitable and just system, then that's not good enough.
My only issue with fixed age-of-consent laws is that they cannot take into consideration the fact that emotional age and intellectual age almost never correspond to physical age. Such laws are understandable only in that neither the other person nor a court can accurately assess emotional and intellectual age, for the most part, making any allowance for the fact that different people mature at different rates wholly unreasonable.
Once a method of testing people's maturity within both hemispheres of the brain on a non-subjective and non-invasive basis has been developed and has itself matured to the point where it is practical and cost-effective for it to be part of a regular check-up, then maybe - just maybe - the laws could be changed to match the reality of the situation for each individual rather than whatever the culture says should be the reality for all people with individuality cut out.
As things stand, three things are certain. Firstly, the law does not protect everyone who needs to be protected. Secondly, the law obstructs people who need no protection and are being denied te right to be themselves. Finally, nothing can be done to fix this, the best any society can do is figure out one or more compromises so as to limit the defectiveness of the law to within reasonable parameters.
I'll also throw in that underage victims of abuse (especially violent or traumatic abuse) deserve every protection and every opportunity for healing that society can afford, whether or not the parents earn money off the children. This is not the case in many parts of the US - in Oregon, many jobs cannot be obtained if a person has been involved in such abuse, even if they were a victim and/or underage, and Oregon is considered one of the more liberal, socially-aware States. The victim, not the abuser, is subject to harassment from society. The legal system isn't much better for victims - only something like 5% of filed sexual abuse complaints result in prosecution, and only a small fraction of those are ever successful, with juries likely to blame the victim for allowing it.
It may be that a better system doesn't lead to even one single additional prosecution or conviction, but the chances are that a better system would lead to the RIGHT 5% of complaints leading to prosection or conviction of the RIGHT perpetrator. As things stand, the abysmally low conviction rate isn't helped by the fact that subsequent scrutiny of evidence handed over, DNA testing, etc, leads to people being found innocent of the crimes for which they are convicted. The chances are that far more genuine cases of abuse take place than there are convictions (whether the specific crime they were convicted of actually happened or not), so it follows that for every person exonerated, there is a perp who has been ignored by the police. Perhaps for that same crime, perhaps for a different one, but the police weren't where they needed to be.
Again, I've no idea how you'd fix that. The whole system, from start to finish, is flawed and defective, but knowing that doesn't make any difference if there's no way of correcting those flaws and defects. We don't have Asimov's "mind-probe" (though fMRI tests on subjects who lied or told the truth would seem to suggest we're getting closer), we don't have any means of conducting extensive psychological testing on the jury pool, the police lack time-travel devices to see what actually happened and to whom, and as perps look just as human as everyone else, they can't be indisputably identified or distinguished by anyone other than a victim - and not always then.
What we have is flawed and morally and ethically repugnant, but short of someone discovering Torchwood in the real world, I don't see any change that will make any real difference happening for many decades - possibly centuries, depending on how much of a difference would be needed to count as "real"
"We've royally buggered things up, software porting to Sugar is limited, quality of code is limited, and developers are leaving, so we'll outsource management of the project to someone who can handle it. Besides, it'll be easier to keep Microsoft happy, if we can deny all responsibility for Sugar working and Windows not."
Dunno, but the ability to turn regions off and protection against jamming/blocking would seem contradictory. If the satellites are capable of directionally shutting down or encrypting, and if the Pentagon's secure network computers are prone to being attacked in cyberwarfare, then the easiest way to jam/block is at source. You don't need to know the encryption, the Pentagon computers can do all that for you, you need only know the command sequence itself. Less, if you know the Americans are going to use the function at a given time for a given location - if viruses and malware can get onto nominally highly secure computers on isolated classified networks (which has been admitted to), then viruses that include a fuzzy-match version of sed can get onto those computers.
Is this at all likely? No, but then neither is jamming equiptment. To jam successfully, a person has to be using an omnidirectional arial, but why would any sane person do that for a signal that HAS to be over a well-determined fraction of the sky? A loop ariel would be serviceable, and would still receive over a far wider angle than you needed, but would make ground-level jamming all but impossible. (A loop ariel loses sensitivity as you approach 90' to the direction of the ariel. So, if the ariel is always level, any jamming device on or very close to the ground will have virtually zero reception.) Now, since the satellites are a fixed angle apart and a fixed number will be above the horizon at any given time, you never need an ariel capable of receiving from a wider angle than (total satellites * angle between each) along one axis. Since all the satellites are at a fixed altitude, to within the maximum accuracy of the signal pulses, you never need to worry about a wider angle than the maximum difference between any two satellites along the second axis. The ariel would need to be able to steer a little, but a simple cog with a release mechanism would suffice to handle the rotation and restore to the original position. Almost completely unjammable except by something directly between the transmitter and receiver, with virtually no extra complexity, and all of that is at the reception end - which is much easier and cheaper to upgrade.
Depending on just how heavy the queries are going to be, and how standards-compliant you can require things to get, you may be able to have the database replicated onto a lighter-weight server. This would have the bonus "selling point" of being faster than the main system, and have the "selling point" at your end that a read-only cache is both a standard and secure method of providing the required service.
OpenSolaris is hybrid open/closed. Solaris was originally closed, so the style and quality control will still be at least partially along closed lines, but is now open so will contain some style and quality control aspects of open source. Without knowing the ratio, however, it is not a particularly useful kernel to study. It is most useful as a sanity check, as it must be somewhere between the two styles, but since you cannot define where between, it has no value in the initial analysis.
This leaves just one closed-source kernel and two open source kernels. If you were to perform a comparitive study between drugs on the basis of three subjects, you'd be laughed out of the research field so fast you'd hit escape velocity and be the first human on Mars.
Now, what OS' could potentially be studied? This depends on who can be talked into supplying software source and whether the researchers are willing to sign NDAs. I'd argue that you want at least three OS' in the closed and open categories, and two in the hybrid - one in each direction.
The closed-source OS' you'd want to examine - if you could get the code - would be Windows (as is already the case), LynxOS-178 (a moderately battle-hardened RTOS) and Trusted IRIX (massively hardened secure OS). In principle, the three design methods and requirements are different to such an extreme that this will give you the best possible picture of the range of code development practices. The idea would be to not test code that would have had similar constraints, because that only tells you about one aspect of software design. If, on the other hand, you pick three points that are more-or-less on different extreme points on the boundary of closed source design, you can get a picture of what that boundary might look like. Substituting any RTOS and Trusted OS for LynxOS-178 and Trusted IRIX respectively would work just as well, the idea is to get the extremes, not get specific implementations of those extremes.
The open-source OS' would include Linux and one of the BSDs - but it has to be a BSD that follows a markedly different ideology, so OpenBSD might be a good choice - but would probably also include one of the BeOS clones and/or a research OS like ExoPC. Again, the idea is to mark out the boundary as carefully as possible. The more data points you have, from points as far apart as possible, the better the understanding of that boundary.
What you will end up with is not a handful of values that have no meaningful basis for comparison, but two overlapping shapes. It'll look something like a Venn diagram. Hybrid methodologies will not merely appear somewhere between the two furthest extremes between the two shapes, but must exist only within the intersection between the two groups, and only there. No other midpoints would be valid.
To test this, we'd use a selection of hybrid designs. Plan9 started closed but has been open the longest, so I'd expect that to be in the intersecion but closer to the midpoint of the open source group than the midpoint of the closd source group. OpenSolaris has been open a relatively short time, so I'd expect the converse to be true for that.
This would still not be a great study, but it would increase the number of data points and provide some measure of control over the experimet. Since you cannot sensibly have a control group in this, the only controls you can assert are those of ensuring the samples are representative of the diversity in techniques, ensuring that the software is correctly classified, and ensuring that the results are sanity-checked.
I'd have liked it if Fresco/Berlin had been able to sustain development. It died a while back, but looked like a serious contender for competing with X11. Of the X11 window managers and environments, I miss things like panning windows - a feature of OLVWM - where a desktop could be larger than the physical screen. Tiling physical screens with a desktop selector just isn't the same, especially when some applications force the windows to be oversized. It's a pain to flip desktops, rather than scroll. Likewise, I miss the Rooms concept, where desktops could themselves contain desktops. Heirarchical systems like that are a clean way of subdividing things.
My main bone of contention with X11 is that it's not being developed seriously as a GUI interface for modern machines. It seems that most of the development is going into code cleanups (important), bugfixes (important) and other maintenance functions. But that's just it - this is all maintenance stuff. The tree needed the reorganization, the code needed to be more modular, etc - nobody is disputing that. On the other hand, threading is overdue and secure X11 channels are insanely overdue. The configuration file changes make things simpler, but it makes it harder to maximise the use of the monitor and graphics cards, even though it's easier (and safer) for the "standard" modes. Simplification is good, but any loss of capability is a regression.
The console is good - and fast - for many tasks, and with the introduction of framebuffers some time back, is capable of many of the tasks people had to use GUIs for in the past. To make the best use of it, though, you really need GNU Screen, and Screen just isn't being maintained that much any more. Really, with framebuffer support and other graphics features for consoles being considered, some of the features of Screen might have to be moved into the kernel in order to function correctly.
I don't use the option of serial-port consoles, so I'm not sure how capable those are these days. PCs are not in the same league as minicomputers or mainframes, so I doubt anybody is looking to hook up a couple of hundred VT220 terminals any time soon, but it is an interface and the underlying code for a terminal is independent of where that terminal is physically located. It should make no difference to Linux whether you are using the local keyboard/screen, a terminal on the end of a serial cable, or indeed a terminal on the end of a USB line.
It could be a capcha-like concept, whereby it's not something that can be automatically joined but requires actual human intervention. It could be a damage limitation exercise - copyright violations can't be indefinite, but are constrained by the access time. Depending on how it's done, it could be a familiarization/educational exercise in teaching students how to distinguish between fair use and copyright infringement. (Bear in mind that "fair use" at an educational facility is often not the same as "fair use" elsewhere by special agreement, because researchers need rather more freedom than the average person and have limited budgets.) It could even be a way to try and encourage students to use P2P for other things than just sharing music and Doctor Who rips. Really, there are so many possible ways something like this could be approached and implemented that assumptions about their reasoning are as futile as trying to stop students filesharing.
There is that, but even within a minor release, 2.6.16 and 2.6.16.60 are probably rather different. There are also kernels with patchsets - 2.6.25 and 2.6.25-mm1 are most certainly different. Then there are the distro modifications. Some will be bugfixes, yes, but there are many drivers, supplementary features and additional architectures out there - some that don't even appear in the -mm patches, as they're too new, too experimental, too unknown, or even too different from the requirements in the kernel standards document. These are the sorts of things distros - especially distros targetting a specific market - are going to include.
Hell, back when I was doing the FOLK patches, I nearly doubled the size of the kernel by including interesting projects. I could easily find 22 megabytes or so of sourcecode that simply wasn't available from any of the well-known kernel repositories. These days, the kernel is a lot bigger, but there are also a lot more people working on things that are frankly fascinating. POHMELFS is unlikely to appear in the vanilla kernel any time soon, but it's an amazing piece of work and a highly novel way of going about things. Very little of Lustre has ever made it into the vanilla kernel and that's a whole lot more orthodox.
I am rather less interested in Roman accounting tallies than I am in differential calculus, group theory and chaos theory. Besides which, would you rather trust Arabs or Alex the Parrot?
Current work on Linux per-process capabilities, role-based access controls and mandatory access controls may render the concept of "root" or a "superuser" under Linux obsolete. What would you need such a user account for? But if there is no superuser, in the traditional sense of the term, then there is no account on the system that would grant the air force (or anyone else) total control of that system. Control would be properly segmented and independently managed, limiting the value of such an attack. Well, it would need to be via the kernel, if no user had those access rights, and it would need to be via a user that could load things into the kernel, and it would need to make use of some exploitable kernel bug that bypassed the security modules.
I think that kind-of goes along with what I was saying. Personally, I'd be skeptical of using a chute to scrub speed on such a device, unless it's kept in a way that is partially opened from the start, so that you could limit or avoid tangling problems. Most modern aircraft have ways of altering the shape of the airfoil to increase air resistance. Although flaps would be impractical on a home-made wing, the basic idea of deforming the wing to create drag is certainly possible. The underling problem is that the wing must increase the lift produced at the same time as slowing down, or you'll stall but at too great a forward velocity to safely crash. A parachute avoids that problem, which is actually quite a difficult one to solve, but creates its own engineering difficulties because you'll have hellish turbulence from the wing and jet(s) - even with the jet(s) switched off. (If the jets are on, the problems mostly centre around the cooking of organic fibres.)
There are programs for evaluating airfoils and there are at least 1,500 published (I guess you could call them open source) airfoils with well-known and well-tested characteristics. It is entirely possible that there is a single wing that would meet all requirements, or that there are two wings where each independently meets the criteria for some part of such a system AND where you could realistically make one behave like, or perhaps be modified into, the other as needed. There are "obvious" one-shot ways of doing something like that. If you could make those cheap enough, that would be sufficient.
With most of the bigger news stories from the Land Formerly Known As Burma being journalists being kicked out, and some of the worst-hit regions in China yet to be reached by anyone other than troops in helicopters, we have exactly two news stories: "something really bad happened in one place, but we'll be shot if we tell you" and "something really bad happened in another place, but it's so bad that we can't get there to tell you". Even if Burma was directly hit by the second cyclone, if you add two complete unknowns together you still only end up with a complete unknown.
Other news stories this month have included finding a statue of Julius Caesar having a midlife crisis, a discovery that pilot whales are bloody fast, and a discovery that the duck-billed platypus is actually an alien from the planet Zark. Oh, and that all of the presidential candidates are loonies, but it's unclear this really qualifies as news - for this or any other campaign.
Ah! This must be why very low-level drops, as used by special forces, are considered bloody dangerous (the chute needs a non-zero time to open), why the rectilinear parachutes are considered more steerable but more prone to entanglement than classical parachutes (which, ergo, means that you can EITHER have predictable steering OR predictable opening, but not both), why the rate of parachute failures on the Thrust-SSC car was unexpectedly high, or why a device invented in 1595 and is intrinsically very simple has a vastly greater mechanical failure rate than far more complex machines that are far newer (and therefore have had far less development time).
In other words, you speak as someone who skydives rather than as someone who is emotionally detached, impartial and observant. Nobody, no matter what the subject, can be impartial when it comes to their own passionate interests. This is not a failing of a person, or a community. It is simply a product of being up close to that degree. If you like, the consequence of being able to see a tree is that you CANNOT see the woods for the trees. The converse is also true. Those distant enough to see the woods CANNOT see the trees. It's an "uncertainty principle" of the way observation works, if you like.
Is landing difficult? Probably, but obviously not impossible - hang-glider enthusiasts land fixed-wing vehicles all the time. Is it more difficult than landing a parachute? Not the problem I'm concerned with, I'm concerned with the mechanical operation of the device, not the mechanical operation of the human, so frankly m'dear, I don't give a damn. Are hang-gliders dangerous? Well, yes, I probably wouldn't want to take off at an altitude so low or where conditions were otherwise so unfavourable that sufficient airspeed to achieve the required level of lift was simply not possible. Now, strap on an engine, so it becomes a microlight, then circumstances change. Microlights are still very dangerous, but not at low altitude. In fact, they'd be quite useless if they couldn't operate from the ground.
Slashdot Mirror
Jade is a semi-precious lump of rock. Whilst UIDs greater than 4 digits are lumps, the relationship to rock (one of the few good music genres) or preciousness (they don't even LOOK like a Ring of Power) is uncertain at this time.
Girls... yes, I do believe I have heard the word used in reference to some members of the population. Never encountered any.
The poster didn't want SQL, so I'm guessing they ruled out SQLite on those grounds. QDBM and Oracle/Sleepycat/Berkley DB-4 would be the smallest, fastest "pure" flat-file databases that are also cross-platform - at least, that I know of.
With the evolution of very low power, high resolution MRI systems, that's entirely possible. Highly sensitive sensors are generally cheaper than 2.5T (and would be a LOT cheaper than 9.2T) magnets and accompanying shielding. On the other hand, any new field is going to take time to move from early uses to mainstream, and getting sensors to the limits of existing MRI technology is going to take a whole lot longer. (It's one thing to demonstrate basic imaging - which is quite good enough for many purposes. It's quite another to get to the point of imaging down to the level of individual neurons. I'm guessing the 12T magnets used in animal experiments provide far more detail yet.)
...there may be some categories of "inoperable" brain tumours that are inoperable because humans have too low a level of precision. Such tumours would be removable by such a method. There have been many advances in tele-surgery since early work in the early 1990s (Surgeons in Russia operated on patients in America, for example) but this is definitely a lot further forward than might have been expected from the pioneering efforts.
But you CAN test a design, and you CAN validate that a given block of code is functionally identical to a formal design for that same block, ergo for some cases you CAN use testing methods to (indirectly) validate locking even when a direct test would not be possible.
is, in fact, just part of a secret government conspiricy. The REAL constitution is in the archives of Torchwood, in Cardiff, and reads: "Nyah! What's up doc?"
I'm more interested in knowing why, when they plotted the path of the chainmail it looked like suit of, well, chainmail. This is a clear indication of the fractal nature of reality, such that our terminology for spam actually infects the upper layers of reality to the point where even couch potatos are zombified into following this ancient topology.
The emotional side of the brain is of critical importance, especially because that is the side of the brain that reacts strongest to pleasure, to hormones and to the senses. Oh, I'm sure there are people - especially amongst the geeks - who can get aroused by a paper on superstring theory or who reach the dizzying heights of ecstacy from filling in a gap in the proof of Poincaire's Conjecture. However, that is probably not very common.
Now, geeks are not known for building up emotional maturity very quickly, and pre-science mindsets don't build up intellectual maturity very quickly. If you required both to exceed the age of consent, the population will plummet. Personally, I couldn't give a damn if the pre-science lot died off, but I'm going to be fair then I guess any two out of the three maturity indexes could be considered sufficient. However, if the only mindset we can easily examine is the intellectual one, then that leaves a lot of combinations that should be legal according to common sense left illegal. If we want to create (or just imagine) a wholly rational, equitable and just system, then that's not good enough.
Once a method of testing people's maturity within both hemispheres of the brain on a non-subjective and non-invasive basis has been developed and has itself matured to the point where it is practical and cost-effective for it to be part of a regular check-up, then maybe - just maybe - the laws could be changed to match the reality of the situation for each individual rather than whatever the culture says should be the reality for all people with individuality cut out.
As things stand, three things are certain. Firstly, the law does not protect everyone who needs to be protected. Secondly, the law obstructs people who need no protection and are being denied te right to be themselves. Finally, nothing can be done to fix this, the best any society can do is figure out one or more compromises so as to limit the defectiveness of the law to within reasonable parameters.
I'll also throw in that underage victims of abuse (especially violent or traumatic abuse) deserve every protection and every opportunity for healing that society can afford, whether or not the parents earn money off the children. This is not the case in many parts of the US - in Oregon, many jobs cannot be obtained if a person has been involved in such abuse, even if they were a victim and/or underage, and Oregon is considered one of the more liberal, socially-aware States. The victim, not the abuser, is subject to harassment from society. The legal system isn't much better for victims - only something like 5% of filed sexual abuse complaints result in prosecution, and only a small fraction of those are ever successful, with juries likely to blame the victim for allowing it.
It may be that a better system doesn't lead to even one single additional prosecution or conviction, but the chances are that a better system would lead to the RIGHT 5% of complaints leading to prosection or conviction of the RIGHT perpetrator. As things stand, the abysmally low conviction rate isn't helped by the fact that subsequent scrutiny of evidence handed over, DNA testing, etc, leads to people being found innocent of the crimes for which they are convicted. The chances are that far more genuine cases of abuse take place than there are convictions (whether the specific crime they were convicted of actually happened or not), so it follows that for every person exonerated, there is a perp who has been ignored by the police. Perhaps for that same crime, perhaps for a different one, but the police weren't where they needed to be.
Again, I've no idea how you'd fix that. The whole system, from start to finish, is flawed and defective, but knowing that doesn't make any difference if there's no way of correcting those flaws and defects. We don't have Asimov's "mind-probe" (though fMRI tests on subjects who lied or told the truth would seem to suggest we're getting closer), we don't have any means of conducting extensive psychological testing on the jury pool, the police lack time-travel devices to see what actually happened and to whom, and as perps look just as human as everyone else, they can't be indisputably identified or distinguished by anyone other than a victim - and not always then.
What we have is flawed and morally and ethically repugnant, but short of someone discovering Torchwood in the real world, I don't see any change that will make any real difference happening for many decades - possibly centuries, depending on how much of a difference would be needed to count as "real"
"We've royally buggered things up, software porting to Sugar is limited, quality of code is limited, and developers are leaving, so we'll outsource management of the project to someone who can handle it. Besides, it'll be easier to keep Microsoft happy, if we can deny all responsibility for Sugar working and Windows not."
Is this at all likely? No, but then neither is jamming equiptment. To jam successfully, a person has to be using an omnidirectional arial, but why would any sane person do that for a signal that HAS to be over a well-determined fraction of the sky? A loop ariel would be serviceable, and would still receive over a far wider angle than you needed, but would make ground-level jamming all but impossible. (A loop ariel loses sensitivity as you approach 90' to the direction of the ariel. So, if the ariel is always level, any jamming device on or very close to the ground will have virtually zero reception.) Now, since the satellites are a fixed angle apart and a fixed number will be above the horizon at any given time, you never need an ariel capable of receiving from a wider angle than (total satellites * angle between each) along one axis. Since all the satellites are at a fixed altitude, to within the maximum accuracy of the signal pulses, you never need to worry about a wider angle than the maximum difference between any two satellites along the second axis. The ariel would need to be able to steer a little, but a simple cog with a release mechanism would suffice to handle the rotation and restore to the original position. Almost completely unjammable except by something directly between the transmitter and receiver, with virtually no extra complexity, and all of that is at the reception end - which is much easier and cheaper to upgrade.
Depending on just how heavy the queries are going to be, and how standards-compliant you can require things to get, you may be able to have the database replicated onto a lighter-weight server. This would have the bonus "selling point" of being faster than the main system, and have the "selling point" at your end that a read-only cache is both a standard and secure method of providing the required service.
This leaves just one closed-source kernel and two open source kernels. If you were to perform a comparitive study between drugs on the basis of three subjects, you'd be laughed out of the research field so fast you'd hit escape velocity and be the first human on Mars.
Now, what OS' could potentially be studied? This depends on who can be talked into supplying software source and whether the researchers are willing to sign NDAs. I'd argue that you want at least three OS' in the closed and open categories, and two in the hybrid - one in each direction.
The closed-source OS' you'd want to examine - if you could get the code - would be Windows (as is already the case), LynxOS-178 (a moderately battle-hardened RTOS) and Trusted IRIX (massively hardened secure OS). In principle, the three design methods and requirements are different to such an extreme that this will give you the best possible picture of the range of code development practices. The idea would be to not test code that would have had similar constraints, because that only tells you about one aspect of software design. If, on the other hand, you pick three points that are more-or-less on different extreme points on the boundary of closed source design, you can get a picture of what that boundary might look like. Substituting any RTOS and Trusted OS for LynxOS-178 and Trusted IRIX respectively would work just as well, the idea is to get the extremes, not get specific implementations of those extremes.
The open-source OS' would include Linux and one of the BSDs - but it has to be a BSD that follows a markedly different ideology, so OpenBSD might be a good choice - but would probably also include one of the BeOS clones and/or a research OS like ExoPC. Again, the idea is to mark out the boundary as carefully as possible. The more data points you have, from points as far apart as possible, the better the understanding of that boundary.
What you will end up with is not a handful of values that have no meaningful basis for comparison, but two overlapping shapes. It'll look something like a Venn diagram. Hybrid methodologies will not merely appear somewhere between the two furthest extremes between the two shapes, but must exist only within the intersection between the two groups, and only there. No other midpoints would be valid.
To test this, we'd use a selection of hybrid designs. Plan9 started closed but has been open the longest, so I'd expect that to be in the intersecion but closer to the midpoint of the open source group than the midpoint of the closd source group. OpenSolaris has been open a relatively short time, so I'd expect the converse to be true for that.
This would still not be a great study, but it would increase the number of data points and provide some measure of control over the experimet. Since you cannot sensibly have a control group in this, the only controls you can assert are those of ensuring the samples are representative of the diversity in techniques, ensuring that the software is correctly classified, and ensuring that the results are sanity-checked.
That just goes to show how bad the latency issues are.
My main bone of contention with X11 is that it's not being developed seriously as a GUI interface for modern machines. It seems that most of the development is going into code cleanups (important), bugfixes (important) and other maintenance functions. But that's just it - this is all maintenance stuff. The tree needed the reorganization, the code needed to be more modular, etc - nobody is disputing that. On the other hand, threading is overdue and secure X11 channels are insanely overdue. The configuration file changes make things simpler, but it makes it harder to maximise the use of the monitor and graphics cards, even though it's easier (and safer) for the "standard" modes. Simplification is good, but any loss of capability is a regression.
The console is good - and fast - for many tasks, and with the introduction of framebuffers some time back, is capable of many of the tasks people had to use GUIs for in the past. To make the best use of it, though, you really need GNU Screen, and Screen just isn't being maintained that much any more. Really, with framebuffer support and other graphics features for consoles being considered, some of the features of Screen might have to be moved into the kernel in order to function correctly.
I don't use the option of serial-port consoles, so I'm not sure how capable those are these days. PCs are not in the same league as minicomputers or mainframes, so I doubt anybody is looking to hook up a couple of hundred VT220 terminals any time soon, but it is an interface and the underlying code for a terminal is independent of where that terminal is physically located. It should make no difference to Linux whether you are using the local keyboard/screen, a terminal on the end of a serial cable, or indeed a terminal on the end of a USB line.
It could be a capcha-like concept, whereby it's not something that can be automatically joined but requires actual human intervention. It could be a damage limitation exercise - copyright violations can't be indefinite, but are constrained by the access time. Depending on how it's done, it could be a familiarization/educational exercise in teaching students how to distinguish between fair use and copyright infringement. (Bear in mind that "fair use" at an educational facility is often not the same as "fair use" elsewhere by special agreement, because researchers need rather more freedom than the average person and have limited budgets.) It could even be a way to try and encourage students to use P2P for other things than just sharing music and Doctor Who rips. Really, there are so many possible ways something like this could be approached and implemented that assumptions about their reasoning are as futile as trying to stop students filesharing.
Hell, back when I was doing the FOLK patches, I nearly doubled the size of the kernel by including interesting projects. I could easily find 22 megabytes or so of sourcecode that simply wasn't available from any of the well-known kernel repositories. These days, the kernel is a lot bigger, but there are also a lot more people working on things that are frankly fascinating. POHMELFS is unlikely to appear in the vanilla kernel any time soon, but it's an amazing piece of work and a highly novel way of going about things. Very little of Lustre has ever made it into the vanilla kernel and that's a whole lot more orthodox.
I am rather less interested in Roman accounting tallies than I am in differential calculus, group theory and chaos theory. Besides which, would you rather trust Arabs or Alex the Parrot?
Current work on Linux per-process capabilities, role-based access controls and mandatory access controls may render the concept of "root" or a "superuser" under Linux obsolete. What would you need such a user account for? But if there is no superuser, in the traditional sense of the term, then there is no account on the system that would grant the air force (or anyone else) total control of that system. Control would be properly segmented and independently managed, limiting the value of such an attack. Well, it would need to be via the kernel, if no user had those access rights, and it would need to be via a user that could load things into the kernel, and it would need to make use of some exploitable kernel bug that bypassed the security modules.
Well, apparently so did the judge, so you're in good company.
There are programs for evaluating airfoils and there are at least 1,500 published (I guess you could call them open source) airfoils with well-known and well-tested characteristics. It is entirely possible that there is a single wing that would meet all requirements, or that there are two wings where each independently meets the criteria for some part of such a system AND where you could realistically make one behave like, or perhaps be modified into, the other as needed. There are "obvious" one-shot ways of doing something like that. If you could make those cheap enough, that would be sufficient.
Other news stories this month have included finding a statue of Julius Caesar having a midlife crisis, a discovery that pilot whales are bloody fast, and a discovery that the duck-billed platypus is actually an alien from the planet Zark. Oh, and that all of the presidential candidates are loonies, but it's unclear this really qualifies as news - for this or any other campaign.
Orange paper plates.
In other words, you speak as someone who skydives rather than as someone who is emotionally detached, impartial and observant. Nobody, no matter what the subject, can be impartial when it comes to their own passionate interests. This is not a failing of a person, or a community. It is simply a product of being up close to that degree. If you like, the consequence of being able to see a tree is that you CANNOT see the woods for the trees. The converse is also true. Those distant enough to see the woods CANNOT see the trees. It's an "uncertainty principle" of the way observation works, if you like.
Is landing difficult? Probably, but obviously not impossible - hang-glider enthusiasts land fixed-wing vehicles all the time. Is it more difficult than landing a parachute? Not the problem I'm concerned with, I'm concerned with the mechanical operation of the device, not the mechanical operation of the human, so frankly m'dear, I don't give a damn. Are hang-gliders dangerous? Well, yes, I probably wouldn't want to take off at an altitude so low or where conditions were otherwise so unfavourable that sufficient airspeed to achieve the required level of lift was simply not possible. Now, strap on an engine, so it becomes a microlight, then circumstances change. Microlights are still very dangerous, but not at low altitude. In fact, they'd be quite useless if they couldn't operate from the ground.