I respect Theo as someone who gets things done, as someone who is great on the frontier of computer security, and as someone who is absolutely essential if software security is to be done right.
I neither respect him OR those who follow him for their attitudes, however. I don't know how long Theo's been in programming, but I believe it likely that I've hacked for longer, better and over a wider range of architectures and programming languages. I've probably worked on a wider range of networking infrastructures, a wider range of Operating Systems and in far more countries than most of the OpenBSD and OpenSSH folk.
Does that give me airs? No. Does that give me the right to question tactice? Oh, certainly. What use is having breadth of knowledge if you never employ it to correct those with depth of specialised knowledge? Specialists are great, nothing wrong with them and you often NEED them, but specialists need generalists in order to make the best use of their skills. Too limited a horizon can make for bad decisions that simply aren't visible to specialists.
A broad horizon, on its own, is equally useless, as you don't get the depth of vision. The ideal is for generalists and specialists to work together, each complimenting the other's skill sets. When that does not happen, the specialist needs to go first, the generalist can then make adjustments, but eventually you'll need to go back to a specialist to progress beyond a certain point.
The FOLK version of OpenSSH is the generalist stage. It will work towards making a more generalized OpenSSH, with a greater range of features, but sooner or later it will need to either re-merge with the classical OpenSSH -or- have a Theo-like person to take over, to drive it to where it needs to go. This is merely a course correction fork.
But the initial drop is indeed there. It's nothing extensive, at the moment - the last snapshot of OpenSSH, with a bunch of patches thrown in, but the fork does exist and it is more than just baseline. I'm calling it openssh-folk, as the FOLK project is specifically for the purpose of severely overloaded software (which is the direction I intend to take this fork).
...to fork OpenSSH, get those patches in and do some decent testing. Hell, why not? You're absolutely right that there is a serious attitude problem going on with the development team - which, IIRC, was the reason OpenSSH was started in the first place. A shitty attitude from SSH.
If OpenSSH's team are worthy enough, then people will stay with them and the fork will fade into history. If, as I suspect, OpenSSH is mostly popular because there are no serious competitors (the rest are infinitely worse), then the moment serious alternatives exist, those alternatives will supplant OpenSSH as the secure system to use.
Anyone can be on top of the heap, when there's no meaningful heap to speak of.
...that a number of patches exist for OpenSSH (speedups, code cleanups, extensions, etc) that aren't getting folded into the baseline. Even if the patches (as they stand) don't meet the coding standards for OpenSSH (there are some?), you really should be seeing efforts to either get the patch writers to reformat to standards OR have core developers recode them.
OpenSSH is limited to IPv4 and IPv6. Limited? Well, yes. Linux supports many non-IP stacks, as do other *nix OS'. So long as you have some component to handle the making of connections and the sending of packets, the rest of OpenSSH doesn't need to care what sort of network you're using or what the transport mechanism is.
I believe OpenSSH can take advantage of some crypto hardware, but I don't recall seeing any announcements that it could use crypto drivers (or crypto functions) in the OS. It links to OpenSSL, but I don't recall seeing any provision for GnuTLS.
Is it the best crypto package out there (SSL included)? Yes. Is it the best it could be? Not by a long shot. Is it the best that it should be, given the code available (both for OpenSSH and as related libraries)? Not even close.
OpenSSH is every bit as "enterprise" as SSH - in fact, for some things, I'd say more so. Does that give the OpenSSH team any excuse to slack off? No - they should be so far ahead, by now, that SSH seems as ancient as the Pyramids and as user-friendly as a unicycle NASCAR.
Of course, we could settle the dispute by bribing^H^H^H^H^H^H^Hlobbying to make IPSec a Federally-mandated standard for all Internet-based computers. Then application-level crypto would cease to be important and we could get onto something useful, like Microsoft-bashing.
Electrons and protons (almost) certainly could not exist in a magnetar. If you collide a proton with an electron, you get a neutron. And at the densities of a magnetar, such collisions would be inevitable. I guess it would be possible to have electrons but NO protons, or protons with NO electrons, but you certainly couldn't have both.
One of the problems with any of the species of giant squid (there are two that are known of) is that they absolutely require a very high pressure to survive. This makes recovery of a live one very difficult. But it could be done.
Start off with a hollow tube. I would suggest a tube about 60' in length (giant squid grow up to 40', and you have to allow time for this to work) and about 10' in diameter. Possibly a bit more. The tube walls need to be somewhere between 10'-20' in thickness and be good-quality steel. Each end needs to have a door that can close and be 100% watertight. Both the door and locking mechanism have to survive pressures of around 400 atmospheres or more. There needs to be a motion detector at each end. If either motion detector registers sustained motion for more than some given length of time, both doors shut the moment motion is no longer detected. (ie: whatever is moving is now fully inside.) You also need to set it that once the doors close, bags on either side are forcibly inflated, so the tube rises to the surface. Once it hits the surface, a radio signals where the tube is.
It's a simple system, the pressure is constant on the inside (so the squid won't be affected) and you could scatter any number of these at the required depth. You then just sit back and wait. Eventually, a squid will be caught. You then tow the tube to the aquarium and lower it into a tank. You then pressurise the tank to 400 atmospheres and open the tube.
(Pressure increases by 1 atmosphere about every 25 feet, so the pressures at 10,000 feet - where Giant Squid roam - will be 400 atmospheres. In practice it may be a little more or a little less, but if you aim for the theoretical pressure, the squid should do just fine.)
This would be implementable by any aquarium (with money) right now. They could have a giant squid within a few months at most, if the tube is baited the way the hooks by these researchers were. There are a few difficulties, though. You'd need 300' thick windows to withstand the pressure. Yes, that's feet. The second problem is that it would be almost impossible to put food into the tank. The third problem is that it would cost a LOT of money to build even one tube, and you'd likely lose most of those you drop into the ocean.
(I'm ignoring the practical difficulties in building a containment system large enough for the squid not to be injured by a high speed impact against the doors when they close, or by impact with the side walls when it tries to turn around.)
...would be to use one password you can remember, for everything. Almost. The key is in that "almost". You have a password calculator, on which you enter your password and the name of the facility you want to access as one long string. The calculator uses a hash function to turn that into a meaningless string. You now have one unique password per machine you want to use, but only one password to actually remember. Nothing is written down and if anyone examines the calculator, all they'll see is a device that does MD5 or SHA1 hashes - they won't be able to actually get any passwords from it.
Furthermore, unless someone DOES obtain the calculator AND knows how you identify the machine, you can tell who you like what the password you remember is. They'd still have to guess the hash function and the salt you're using. And if the user doesn't know how the calculator works (only that it does), social engineering won't help in getting the function, even if the cracker got all the other data. A cracker would need to actually try different hashing functions to be able to crack passwords for other sites, which increases the odds of them being detected.
If you need High Availability (ie: almost, but not quite, 100% uptime) then you want two or more boxes which you can either load-balance between (dropping crashed servers from the list) OR which you can fail-over to in the event of a problem.
Use a service monitor, such as Hobbit or Big Brother, to monitor the services. If a service fails once, have it auto-restarted. If it fails repeatedly, have the monitor reboot the box automatically. If the server keeps crashing, or if recovery locks up, then have it notify you to intervene.
If you need guaranteed 100% uptime -and- the software is more reliable than the hardware (for whatever reason), your best bet is to run two boxes in parallel and have BOTH serve all requests, but have your router filter out the latter of any two identical packets. Then, if a box crashes, the other already has the connection running and no fail-over is required. When the crashed box is restored, you then have to replicate the state, but you can afford to take more time over that, as the user won't be aware of it.
If you need even higher levels of availability, then you'd want to move the disks onto a SAN and mirror the disk access - this time, filtering out duplicates in both directions. That way, either computer can crash AND either disk RAID can crash, and you STILL have a functioning system.
You can keep parallelizing, adding redundancy (such as mixing hot-swap and cold-standby), etc, as far as you like for the reliability you require. Need better WAN network reliability? Get two providers and set up dynamic routing over BGP. Let BGP take care of monitoring connectivity and dropping routes that aren't working. That's what routers are designed to do. Ideally, you'd co-locate, have backup IP providers for each site, then replicate transactions on write. If a site totally crashes, you can't avoid the time to fail over, but you can keep it to an absolute minimum.
If the LAN is potentially a weakpoint, then have each server with a line to each WAN router. Have an independent cable running between servers, and use it specifically for replicating states when a server is restarted.
These options range from costing virtually nothing (Big Brother is free) to tens of thousands of dollars or more, depending on the scale of redundancy you want, and give you from a few seconds of downtime perhaps every few months through to no user-visible downtime ever (short of a nuclear attack on all locations you co-locate between).
The Linux Test Project, Web100 (which profiles the network stack) and the TAHI IPv6/IPSec Test Suite should be usable to give you a starting point for validating the kernel. HOL may be usable as a component in formal proofs of components with well-defined behaviour (such as busses, networks, etc). Both TAU and the Performance Application Programming Interface would allow you to create profiles of kernel components running in User-Mode Linux, so allowing developers to spot the causes of things like latency.
These wouldn't solve ALL problems, or even the majority of them, but they would solve some and they would make life easier on developers in the long-run. Are these being used? Well, a glance at the Freshmeat graphs for Web100 shows that it is getting downloaded. This doesn't mean it is getting used, though. The same is true of virtually all of the other code I've mentioned. People have copies, but if the code being submitted is flakey and taking a long time to fix, then maybe the code is not being used as much as it could/should be.
The tools exist, the tools exist on people's hard drives, but unless the tools are being used in practice, that's not going to do any good.
That would be as stupid as handing Haliburton a no-compete contract to rebuild New Orleans, after the over-priced, under-performing fiascos they've been involved with in Iraq. Oh. they did.
Gravitational rounding (as opposed to rounding by any other means) has been put forward as an idea, from what I've heard.
The definition I'd use would be: An object that has gravitational rounding, has as the center of its primary orbit the sun (or, for complex cases, orbits no more planets than suns*), and has non-uniform, non-random composition with a single core.
*This is a cheat. I was going to say "orbits at least one sun" to take into account binary star systems where the center of the planet's orbit may not actually lie inside either sun, and also binary systems where a massive gas giant takes the place of one of the suns, as far as orbits are concerned. However, that doesn't allow for extra-solar planets with no suns at all. It also doesn't allow for a case where you've a moon that orbits two or more planets, but may occasionally also orbit a sun. The Universe is complex, so complex cases will arise, so we need a definition that won't need modifying much later on.
My idea here is to approach this from two directions. The first is to look at all of the bodies in the solar system that are definitely NOT planets, and identify what it is about them that gives them this non-planetness. For example, comets are not planets. Comets have multiple rocky cores - gravity has been insufficient to combine the object into a single mass, it has merely been enough for adhesion to take place. All planets have a single core, including the gas giants. Ergo, this allows us to uniquely define comets and thus filter them out of the bodies of the solar system.
Asteroids are the next ones. It's very likely that most don't have a core at all, but some may. If they do, it'll be a single core. So, we have to look for something else to exclude these. Asteroids are very uniform in composition, as far as anyone can tell, and meteorite fragments that have reached Earth from the asteroid belt would indicate only a very limited subset of elements being present. My guess is that the accretion disk started fairly uniform but eventually acted as a gravitational centrifuge, separating out the elements by mass. The asteroids probably formed from whatever had the right characteristics to be at that distance from the centrifuge.
In comparison, Earth has everything from the heaviest to the lightest elements, in roughly the proportions you'd expect from observation of the table of the abundances of the elements. This would seem to be the case for Mars, Venus, Jupiter and Saturn. We don't know enough about the other planets to be certain, but I'd be surprised if it was any different. So, composition seems to distinguish planets from asteroids.
The problem with too simple a definition is we're going to be forever updating it, then arguing over what should keep the title and what should be downgraded/upgraded. Only, as time goes on, there will be more at stake - politically, economically, etc. Right now, we've only got a few egos and some textbooks to contend with, but that won't be true forever, so we want a definition that will need as little changing as humanly possible.
I would personally opt for PNG for images, to avoid loss of data. Video almost has to be MPEG, as neither MNG nor APNG have really gone anywhere at this time and the BBC's high definition format isn't getting much adoption yet either. For audio, MP4 would seem the best choice - less loss of data, but more likely to be readable in the far future than Ogg Vorbis (which is a shame) or AIFF (yay! AIFF's gonna die!)
No matter what form you store the data in, if you want it readable in the far future, you've got to remember two things - there's no guarantee ANY specific technology will exist, and there's no guarantee ANY specific timeframe for the reading to take place.
What you want, then, is to do the reverse of the language decoding that has taken place over the years. Imagine yourself faced with a puzzle every bit as baffling as Egyptian Hyroglyphics, only stored at a vastly greater information density and probably in an electronic format. What would you want/need, to be able to recover the data?
Well, there would seem to be a few things that are essential. First, the explorer in the future will need to know the data is there and in what form. So, if you're using optical storage, make that clear (along with frequency). If you're using N-state logic, make it clear what N is. If there are M layers, tell them the value of M. You don't need to know all of the technical information, because all they need is where to start looking.
Secondly, the information needs to be correctly indexed. Languages are broken because types of information can be grouped and identified. The same will be true here. So, produce a contents list with corresponding data formats and/or MIME types, along with the offsets within the medium.
Thirdly, a key is a REALLY good idea - something analogous to the Rosetta Stone. Let's say you're using binary logic and a fairly rudimentary FS on the storage medium with text-based directories. The key would be a printout of the root directory in binary, again in ASCII and a third time as a set of records describing the logical layout. The printout would also need the offset of the directory. From this, it would be trivial for someone in the year 3000 to determine how offsets were calculated, how the data was laid on the disk and how the data is connected.
If physical storage is going to be used, ensure the various media used will last about the same length of time. So, if you're aiming for a hundred years, CDs may just about work. But you must NOT have the CD in contact with sulphides or anything else which will destroy the surface. The CD must be kept cold (but not so cold it is damaged) to slow decomposition. It should also be kept somewhere where accidental exposure to UV is impossible.
If you're keeping paper notes with the data, as I've suggested, the paper must be acid-free and the inks must be long-lasting. Most modern paper is of very low grade, as are most modern inks.
If you're looking more at a time capsule that is for the FAR future (we're talking something that happens AFTER Star Trek), then you've got to be extra careful but it should still be possible. I see no reason why you couldn't have physical storage under ideal conditions which could be retrievable after a thousand years or so. You just have to be very careful on what you choose to use. Same with paper. If you're looking to produce the next Beowulf (no, not the clustering technology), then you're probably going to want to look at vellum or some other extremely high-quality medium. I'd also look up early inks on the Internet and modify a recipe that could be used as a refill for a printer ink cartridge. Many early inks are highly stable (iron oxide is one example) and fade more by damage to the medium than decay of the ink.
It was my understanding that American police were immune from personal lawsuits involving actions when they were in uniform and on duty. It's this immunity that's usually been quoted as preventing unlawful death or personal injury lawsuits that are brought through civil law.
There were two underlying causes for the Government to ban travelling together with a common purpose. The first was repeated attempts by latter-day druids to "invade" Stonehenge for the summer solstice. The second was a British version of Woodstock, when 10,000 or so people gathered for a three-day festival on fallow land. It was decided that the only way to stop such people was to ban all travelling involving groups of people.
(I'm trying to figure out who in the House of Commons was the re-incarnation of King Canute...)
White, middle-class people wear clothes from Burtons or Marks & Sparks. Gap is for people who want to seem trendy so badly that they'll wear clothes with a designer label larger than the outfit just so people know what they have. That means it'll be restricted to Yuppies (who have no class) and people who are both idiots -and- of non-European descent (one or the other doesn't count).
They'd still believe it couldn't happen in the US, even if it happened right in front of them. Even if it happened to them, they'd just go to court, perhaps win some compensation, and then still insist it couldn't happen in the US. The only exceptions are those who are convinced that everyone's out to get them, which means they not only lack credibility when it's true, but they'd also fail to identify those things which were important. If you're convinced it either never happens to anyone OR it always happens to everyone, you're simply not going to be capable of observation and critical thinking.
Pirates in Cornwall used tunnels all the time. So did the wreckers and the smugglers. Actually, the wreckers were the worst menace - they'd create fake lighthouses and use them to steer ships onto the rocks. They'd then just wait on the shoreline for wreckage from the cargo holds to come ashore. These days, they're mostly working as advertising execs, American lawyers and SCO advisers.
Under Common Law in the UK, you generally can't be prosecuted for doing anything that any reasonable person might do. This was used very successfully by a man claiming to be the reincarnation of King Arthur, who had been prosecuted along with some of his "knights of the round table" for 'travelling together with a common purpose' - an action prohibited under the Criminal Justice Act. The Law Lords considered his defence that he couldn't go on quests under the CJA a better argument than the Government's.
The TPA is supposed to have safeguards, preventing wanton abuse by the police - otherwise they'd just call everyone a terrorist and sort out who was what over the week they get before having to present some evidence. The police can't just arrest anyone they happen to feel like. Well, they can, they're just going to get bollicked by the courts if they try, as happened in the aforementioned case.
This is no different from in the US, where anyone can physically be arrested by the police and subject to whatever searches the police feel like. The Constitution is just paper, it can't physically intervene. All a person can do is plead their case in court and hope for a sensible judge. (More than a few convictions in the US have been overturned on appeal, because the Miranda rights were violated - deomonstrating that it can take several rounds before anyone pays attention.)
It's also important to note we don't know ALL of the facts of the case. For all we know, British Intelligence may have tipped the police off that an attack was likely on that route, sometime soon. In which case, you're dealing with an entirely different scenario to one where the police were acting spontaneously, without due cause. All we can do, at this point, is guess as to the motives involved and the information the police had posessed. (I shall refrain from drawing inferences about the demonic nature of anti-terror squads that posess.)
...that you're right. I think one thing that makes me nervous is things like the "smart" markup tried be some companies on browsers that have added hypertext to pages. (Microsoft did that, for a while, IIRC.) I don't recall lawsuits over that, even though the pages were functionally different, not merely logically different. It disturbs me that so many thought it a good feature and that the biggest reason it never happened was that the IE team had largely moved on by then.
Slashdot Mirror
I neither respect him OR those who follow him for their attitudes, however. I don't know how long Theo's been in programming, but I believe it likely that I've hacked for longer, better and over a wider range of architectures and programming languages. I've probably worked on a wider range of networking infrastructures, a wider range of Operating Systems and in far more countries than most of the OpenBSD and OpenSSH folk.
Does that give me airs? No. Does that give me the right to question tactice? Oh, certainly. What use is having breadth of knowledge if you never employ it to correct those with depth of specialised knowledge? Specialists are great, nothing wrong with them and you often NEED them, but specialists need generalists in order to make the best use of their skills. Too limited a horizon can make for bad decisions that simply aren't visible to specialists.
A broad horizon, on its own, is equally useless, as you don't get the depth of vision. The ideal is for generalists and specialists to work together, each complimenting the other's skill sets. When that does not happen, the specialist needs to go first, the generalist can then make adjustments, but eventually you'll need to go back to a specialist to progress beyond a certain point.
The FOLK version of OpenSSH is the generalist stage. It will work towards making a more generalized OpenSSH, with a greater range of features, but sooner or later it will need to either re-merge with the classical OpenSSH -or- have a Theo-like person to take over, to drive it to where it needs to go. This is merely a course correction fork.
But the initial drop is indeed there. It's nothing extensive, at the moment - the last snapshot of OpenSSH, with a bunch of patches thrown in, but the fork does exist and it is more than just baseline. I'm calling it openssh-folk, as the FOLK project is specifically for the purpose of severely overloaded software (which is the direction I intend to take this fork).
If OpenSSH's team are worthy enough, then people will stay with them and the fork will fade into history. If, as I suspect, OpenSSH is mostly popular because there are no serious competitors (the rest are infinitely worse), then the moment serious alternatives exist, those alternatives will supplant OpenSSH as the secure system to use.
Anyone can be on top of the heap, when there's no meaningful heap to speak of.
OpenSSH is limited to IPv4 and IPv6. Limited? Well, yes. Linux supports many non-IP stacks, as do other *nix OS'. So long as you have some component to handle the making of connections and the sending of packets, the rest of OpenSSH doesn't need to care what sort of network you're using or what the transport mechanism is.
I believe OpenSSH can take advantage of some crypto hardware, but I don't recall seeing any announcements that it could use crypto drivers (or crypto functions) in the OS. It links to OpenSSL, but I don't recall seeing any provision for GnuTLS.
Is it the best crypto package out there (SSL included)? Yes. Is it the best it could be? Not by a long shot. Is it the best that it should be, given the code available (both for OpenSSH and as related libraries)? Not even close.
OpenSSH is every bit as "enterprise" as SSH - in fact, for some things, I'd say more so. Does that give the OpenSSH team any excuse to slack off? No - they should be so far ahead, by now, that SSH seems as ancient as the Pyramids and as user-friendly as a unicycle NASCAR.
Of course, we could settle the dispute by bribing^H^H^H^H^H^H^Hlobbying to make IPSec a Federally-mandated standard for all Internet-based computers. Then application-level crypto would cease to be important and we could get onto something useful, like Microsoft-bashing.
Electrons and protons (almost) certainly could not exist in a magnetar. If you collide a proton with an electron, you get a neutron. And at the densities of a magnetar, such collisions would be inevitable. I guess it would be possible to have electrons but NO protons, or protons with NO electrons, but you certainly couldn't have both.
Open Groupware seems to be a good alternative, these days - especially if you have your own mail server, as they don't supply one!
Start off with a hollow tube. I would suggest a tube about 60' in length (giant squid grow up to 40', and you have to allow time for this to work) and about 10' in diameter. Possibly a bit more. The tube walls need to be somewhere between 10'-20' in thickness and be good-quality steel. Each end needs to have a door that can close and be 100% watertight. Both the door and locking mechanism have to survive pressures of around 400 atmospheres or more. There needs to be a motion detector at each end. If either motion detector registers sustained motion for more than some given length of time, both doors shut the moment motion is no longer detected. (ie: whatever is moving is now fully inside.) You also need to set it that once the doors close, bags on either side are forcibly inflated, so the tube rises to the surface. Once it hits the surface, a radio signals where the tube is.
It's a simple system, the pressure is constant on the inside (so the squid won't be affected) and you could scatter any number of these at the required depth. You then just sit back and wait. Eventually, a squid will be caught. You then tow the tube to the aquarium and lower it into a tank. You then pressurise the tank to 400 atmospheres and open the tube.
(Pressure increases by 1 atmosphere about every 25 feet, so the pressures at 10,000 feet - where Giant Squid roam - will be 400 atmospheres. In practice it may be a little more or a little less, but if you aim for the theoretical pressure, the squid should do just fine.)
This would be implementable by any aquarium (with money) right now. They could have a giant squid within a few months at most, if the tube is baited the way the hooks by these researchers were. There are a few difficulties, though. You'd need 300' thick windows to withstand the pressure. Yes, that's feet. The second problem is that it would be almost impossible to put food into the tank. The third problem is that it would cost a LOT of money to build even one tube, and you'd likely lose most of those you drop into the ocean.
(I'm ignoring the practical difficulties in building a containment system large enough for the squid not to be injured by a high speed impact against the doors when they close, or by impact with the side walls when it tries to turn around.)
Giant squid (and giant octopi) have amonium-based nervous systems. You'd need to find a good way to prepare it.
Furthermore, unless someone DOES obtain the calculator AND knows how you identify the machine, you can tell who you like what the password you remember is. They'd still have to guess the hash function and the salt you're using. And if the user doesn't know how the calculator works (only that it does), social engineering won't help in getting the function, even if the cracker got all the other data. A cracker would need to actually try different hashing functions to be able to crack passwords for other sites, which increases the odds of them being detected.
Use a service monitor, such as Hobbit or Big Brother, to monitor the services. If a service fails once, have it auto-restarted. If it fails repeatedly, have the monitor reboot the box automatically. If the server keeps crashing, or if recovery locks up, then have it notify you to intervene.
If you need guaranteed 100% uptime -and- the software is more reliable than the hardware (for whatever reason), your best bet is to run two boxes in parallel and have BOTH serve all requests, but have your router filter out the latter of any two identical packets. Then, if a box crashes, the other already has the connection running and no fail-over is required. When the crashed box is restored, you then have to replicate the state, but you can afford to take more time over that, as the user won't be aware of it.
If you need even higher levels of availability, then you'd want to move the disks onto a SAN and mirror the disk access - this time, filtering out duplicates in both directions. That way, either computer can crash AND either disk RAID can crash, and you STILL have a functioning system.
You can keep parallelizing, adding redundancy (such as mixing hot-swap and cold-standby), etc, as far as you like for the reliability you require. Need better WAN network reliability? Get two providers and set up dynamic routing over BGP. Let BGP take care of monitoring connectivity and dropping routes that aren't working. That's what routers are designed to do. Ideally, you'd co-locate, have backup IP providers for each site, then replicate transactions on write. If a site totally crashes, you can't avoid the time to fail over, but you can keep it to an absolute minimum.
If the LAN is potentially a weakpoint, then have each server with a line to each WAN router. Have an independent cable running between servers, and use it specifically for replicating states when a server is restarted.
These options range from costing virtually nothing (Big Brother is free) to tens of thousands of dollars or more, depending on the scale of redundancy you want, and give you from a few seconds of downtime perhaps every few months through to no user-visible downtime ever (short of a nuclear attack on all locations you co-locate between).
It used its attitude control on the phone company.
These wouldn't solve ALL problems, or even the majority of them, but they would solve some and they would make life easier on developers in the long-run. Are these being used? Well, a glance at the Freshmeat graphs for Web100 shows that it is getting downloaded. This doesn't mean it is getting used, though. The same is true of virtually all of the other code I've mentioned. People have copies, but if the code being submitted is flakey and taking a long time to fix, then maybe the code is not being used as much as it could/should be.
The tools exist, the tools exist on people's hard drives, but unless the tools are being used in practice, that's not going to do any good.
That would be as stupid as handing Haliburton a no-compete contract to rebuild New Orleans, after the over-priced, under-performing fiascos they've been involved with in Iraq. Oh. they did.
The definition I'd use would be: An object that has gravitational rounding, has as the center of its primary orbit the sun (or, for complex cases, orbits no more planets than suns*), and has non-uniform, non-random composition with a single core.
*This is a cheat. I was going to say "orbits at least one sun" to take into account binary star systems where the center of the planet's orbit may not actually lie inside either sun, and also binary systems where a massive gas giant takes the place of one of the suns, as far as orbits are concerned. However, that doesn't allow for extra-solar planets with no suns at all. It also doesn't allow for a case where you've a moon that orbits two or more planets, but may occasionally also orbit a sun. The Universe is complex, so complex cases will arise, so we need a definition that won't need modifying much later on.
My idea here is to approach this from two directions. The first is to look at all of the bodies in the solar system that are definitely NOT planets, and identify what it is about them that gives them this non-planetness. For example, comets are not planets. Comets have multiple rocky cores - gravity has been insufficient to combine the object into a single mass, it has merely been enough for adhesion to take place. All planets have a single core, including the gas giants. Ergo, this allows us to uniquely define comets and thus filter them out of the bodies of the solar system.
Asteroids are the next ones. It's very likely that most don't have a core at all, but some may. If they do, it'll be a single core. So, we have to look for something else to exclude these. Asteroids are very uniform in composition, as far as anyone can tell, and meteorite fragments that have reached Earth from the asteroid belt would indicate only a very limited subset of elements being present. My guess is that the accretion disk started fairly uniform but eventually acted as a gravitational centrifuge, separating out the elements by mass. The asteroids probably formed from whatever had the right characteristics to be at that distance from the centrifuge.
In comparison, Earth has everything from the heaviest to the lightest elements, in roughly the proportions you'd expect from observation of the table of the abundances of the elements. This would seem to be the case for Mars, Venus, Jupiter and Saturn. We don't know enough about the other planets to be certain, but I'd be surprised if it was any different. So, composition seems to distinguish planets from asteroids.
The problem with too simple a definition is we're going to be forever updating it, then arguing over what should keep the title and what should be downgraded/upgraded. Only, as time goes on, there will be more at stake - politically, economically, etc. Right now, we've only got a few egos and some textbooks to contend with, but that won't be true forever, so we want a definition that will need as little changing as humanly possible.
No matter what form you store the data in, if you want it readable in the far future, you've got to remember two things - there's no guarantee ANY specific technology will exist, and there's no guarantee ANY specific timeframe for the reading to take place.
What you want, then, is to do the reverse of the language decoding that has taken place over the years. Imagine yourself faced with a puzzle every bit as baffling as Egyptian Hyroglyphics, only stored at a vastly greater information density and probably in an electronic format. What would you want/need, to be able to recover the data?
Well, there would seem to be a few things that are essential. First, the explorer in the future will need to know the data is there and in what form. So, if you're using optical storage, make that clear (along with frequency). If you're using N-state logic, make it clear what N is. If there are M layers, tell them the value of M. You don't need to know all of the technical information, because all they need is where to start looking.
Secondly, the information needs to be correctly indexed. Languages are broken because types of information can be grouped and identified. The same will be true here. So, produce a contents list with corresponding data formats and/or MIME types, along with the offsets within the medium.
Thirdly, a key is a REALLY good idea - something analogous to the Rosetta Stone. Let's say you're using binary logic and a fairly rudimentary FS on the storage medium with text-based directories. The key would be a printout of the root directory in binary, again in ASCII and a third time as a set of records describing the logical layout. The printout would also need the offset of the directory. From this, it would be trivial for someone in the year 3000 to determine how offsets were calculated, how the data was laid on the disk and how the data is connected.
If physical storage is going to be used, ensure the various media used will last about the same length of time. So, if you're aiming for a hundred years, CDs may just about work. But you must NOT have the CD in contact with sulphides or anything else which will destroy the surface. The CD must be kept cold (but not so cold it is damaged) to slow decomposition. It should also be kept somewhere where accidental exposure to UV is impossible.
If you're keeping paper notes with the data, as I've suggested, the paper must be acid-free and the inks must be long-lasting. Most modern paper is of very low grade, as are most modern inks.
If you're looking more at a time capsule that is for the FAR future (we're talking something that happens AFTER Star Trek), then you've got to be extra careful but it should still be possible. I see no reason why you couldn't have physical storage under ideal conditions which could be retrievable after a thousand years or so. You just have to be very careful on what you choose to use. Same with paper. If you're looking to produce the next Beowulf (no, not the clustering technology), then you're probably going to want to look at vellum or some other extremely high-quality medium. I'd also look up early inks on the Internet and modify a recipe that could be used as a refill for a printer ink cartridge. Many early inks are highly stable (iron oxide is one example) and fade more by damage to the medium than decay of the ink.
But 127.0.0.2 has been converted to an airbase and 127.22.91.7 is on active service in Iraq.
If they're not allowed to sue, they'll just have to sue very quietly.
It was my understanding that American police were immune from personal lawsuits involving actions when they were in uniform and on duty. It's this immunity that's usually been quoted as preventing unlawful death or personal injury lawsuits that are brought through civil law.
(I'm trying to figure out who in the House of Commons was the re-incarnation of King Canute...)
White, middle-class people wear clothes from Burtons or Marks & Sparks. Gap is for people who want to seem trendy so badly that they'll wear clothes with a designer label larger than the outfit just so people know what they have. That means it'll be restricted to Yuppies (who have no class) and people who are both idiots -and- of non-European descent (one or the other doesn't count).
They'd still believe it couldn't happen in the US, even if it happened right in front of them. Even if it happened to them, they'd just go to court, perhaps win some compensation, and then still insist it couldn't happen in the US. The only exceptions are those who are convinced that everyone's out to get them, which means they not only lack credibility when it's true, but they'd also fail to identify those things which were important. If you're convinced it either never happens to anyone OR it always happens to everyone, you're simply not going to be capable of observation and critical thinking.
Pirates in Cornwall used tunnels all the time. So did the wreckers and the smugglers. Actually, the wreckers were the worst menace - they'd create fake lighthouses and use them to steer ships onto the rocks. They'd then just wait on the shoreline for wreckage from the cargo holds to come ashore. These days, they're mostly working as advertising execs, American lawyers and SCO advisers.
...would be illegal in Speaker's Corner in Hyde Park. Or, for that matter, even remotely unusual...
The TPA is supposed to have safeguards, preventing wanton abuse by the police - otherwise they'd just call everyone a terrorist and sort out who was what over the week they get before having to present some evidence. The police can't just arrest anyone they happen to feel like. Well, they can, they're just going to get bollicked by the courts if they try, as happened in the aforementioned case.
This is no different from in the US, where anyone can physically be arrested by the police and subject to whatever searches the police feel like. The Constitution is just paper, it can't physically intervene. All a person can do is plead their case in court and hope for a sensible judge. (More than a few convictions in the US have been overturned on appeal, because the Miranda rights were violated - deomonstrating that it can take several rounds before anyone pays attention.)
It's also important to note we don't know ALL of the facts of the case. For all we know, British Intelligence may have tipped the police off that an attack was likely on that route, sometime soon. In which case, you're dealing with an entirely different scenario to one where the police were acting spontaneously, without due cause. All we can do, at this point, is guess as to the motives involved and the information the police had posessed. (I shall refrain from drawing inferences about the demonic nature of anti-terror squads that posess.)
...that you're right. I think one thing that makes me nervous is things like the "smart" markup tried be some companies on browsers that have added hypertext to pages. (Microsoft did that, for a while, IIRC.) I don't recall lawsuits over that, even though the pages were functionally different, not merely logically different. It disturbs me that so many thought it a good feature and that the biggest reason it never happened was that the IE team had largely moved on by then.