That would seem obvious, provided it is fine-grained enough. (Someone can be a dolt in subject X and a genius in subject Y.) As I said in my earlier post, it also needs room for flexibility, as some people naturally work better with those of a different level of ability, so you have to take into account not only an individual's abilities in isolation, but also within some sort of context. (For example, there are likely less able students who actually need to rise to a challenge to get themselves motivated. They need to be streamed higher than their individual scores would suggest in isolation.)
The other thing to consider is that it isn't easy to determine ability in a subject. The best mechanism I can think of is to perform per-subject tests at regularly-spaced intervals, where the test has no upper limit on scores, weighted by the stream they are currently in. The closer to the expected weighted value for that student, the better the fit between stream and student. A score above expectations means the student is quite capable of learning faster. A score below expectations means either a lack of motivation or a lack of ability at that pace. That cannot easily be resolved by tests and would be better resolved by working with the student. A lack of motivation through boredom should result in a faster stream. A lack of ability should mean a slower stream. Failure to communicate well with the teacher should mean going to the nearest stream either side in which communication is improved.
Because this is a massively multi-variable problem, an ideal fit is unlikely for all students. The best anyone can hope for is a superior fit for the majority of students, with no worse than "standard" methods for all others.
Actually, I'm not a teacher, I'm a software engineer. Now, my younger sister teaches molecular biology, my father was a (UK) professor of chemistry and my grandfather was a maths lecturer. Hmmm. Nope, no languages in there. Now, whilst it is certainly true that almost all subjects require brainpower, they exercise different sections of the brain and are therefore not equal in their impact. You cannot simply substitute one thing for another and expect the same results. In general, schools that provide so-called "Classical Education" do appear to produce a disproportionately large number of students with a high level of intellectual ability. This is not to say they are the best - they are actually quite feeble in many areas that "modern" education has mastered. The problem is to identify those elements of "Classical education" that are either inferior or missing from more modern methods that would account for the higher potential. Research in Europe has generally concluded that the element of interest is language. Polyglots have better minds.
I am not interested in qualitative studies for the most part and I have little regard for statistical studies. These are too easily rigged to suit people's preconceptions. Empirical research - where possible - where a definite pathway A-to-B-to-C can be demonstrated is the ideal. In education, empirical research is extremely difficult and it's only recently that technology such as fMRI even existed. Certainly, schools aren't equipt with such technology to routinely scan the kids' brains to determine how the school culture is influencing things. Virtually all other research is culturally biased and heavily statistical with no statement on the level of confidence used in the statistical tests, or indeed what statistical method was used, and how it was ensured that a genuinely representative random sample was obtained.
The understanding of the importance of language comes from a mix of research - some involving fMRI (which lends credibility to it), but most involving qualitative assessments of education and the impact on society from schools in Europe over the past 2,500 years. The amount of data is fairly impressive and there is therefore some credibility beyond the "hard science" to the argument that multi-cultural and linguistically sophisticated education is far superior to insular mono-cultures such as those found in many parts of the US.
I'll finish up with costs. I hold to the belief that skilled work generally produces more than it costs, whereas unskilled work is merely necessary to get any work done at all. If you eliminate unskilled labour as much as possible, transferring it to machines or whatever, and raise the educational standards across the board, the net value of the work done will rise. Since we pay taxes as a percent of our income, and corporations pay taxes according to their earnings, etc, the net value of taxes must also rise, if the populace is adequately educated. Now, the law of diminishing returns does come into play here. You can't improve education forever and expect the wealth generated to go to infinity. That won't happen. What will happen is that it will tend to some upper limit. There is therefore some upper bound where further investment will have no significant benefit. However, investment BELOW that point will generate a substandard return and investing further will reap enough of a return to be profitable all round.
(It's not a simple relationship. Greater education produces not only greater skills but better research, which means that the cost for R&D will fall, with respect to the value of the products developed. This will directly benefit the companies, but because you now have more money circulating, you also have more money collected in taxes. R&D is much more random, however, and therefore it is much harder to predict the impact of higher-quality work. This is something you can really only try and see.)
Total investment in education is as much a myth as the total employmen
Sadly, you are correct. Many kids who are a "problem" are bored. Some because they're smart enough to plough through the material, others because they run into a problem, get stuck, and have to wait 10-15 minutes before the teacher can help. Assuming the teacher ever does, of course.
Based on my own volunteer work in school programs, I would say that class sizes should rarely be above 15-20 in total, and should have 1 teacher/assistant competent in the subject for every 5-7 students. I also think kids should be streamed per subject, with some flexibility for when certain groups of kids happen to work well together. (No, that does not mean cribbing the notes.)
The problem with the existing system is that it is geared around people learning as and when the teacher gets round to it, rather than pushing people as far and as fast as they are able. It is no wonder that kids use drugs, but my guess is that its more to zone out the inadequacies of the educational system as it is to improve learning. You can't accelerate much beyond the speed the material is taught.
Based on research that has been caried out, I think that I'd extend this basic concept by throwing in a second or even a third language, as it appears that the complexity of language is such that learning new languages young boosts the growth of neural connections and seems to improve the capacity to learn. Languages, therefore, may provide a safe alternative to these drugs in that they'll boost intelligence and have no risk of later side-effects.
If OpenMOSIX is compiled into the kernel, the total effort required to set up a Linux cluster is virtually nil - you need to tell it what nodes there are in the cluster and it transparently takes care of the rest. A home user with ZERO programming experience but has two or more computers, a hub and a working knowledge of what an IP address is can configure a rudimentary cluster in under five minutes. It may not be optimal, it'll rely totally on OpenMOSIX to do the process migration, and without any apps that can take advantage of it, it would be a little pointless, but it could be done. It requires no expert knowledge or significant intelligence. If you can operate VI, you can operate a cluster at that level.
Difficulty, therefore, is NOT a significant factor in all of this. Ok, what about expense? Well, you're right that Linux is free. So is OpenMOSIX, OpenMPI (and many other MPI implementations), PVM (another messaging library), Lustre (a very high-performance network file system), many scientific and mathematical applications for clusters, etc. There are clustering patches for PoVRay, and it's always possible to write a script to have multiple machines render parts of images anyway. I'm sure there are other applications out there that I'm not thinking of right now, and it's only a matter of time before more "mundane" applications can take advantage of clustered environments. They already do, on Plan 9, to some degree. Oh, Plan 9 is also free.
Cost would appear not to be a major problem either, then. Optimizing is the only thing that is in any way difficult, and a GUI system that doesn't let you get to the really fine detail won't help there. More time, effort and money is spent on optimizing than on anything else, and I simply can't see any possible way that an OS that is designed for ease-of-use by hiding the intricacies can in any way help in that.
Microsoft delivered the keynote speech for the Supercomputer conference SC|05 in Seattle last year and had a huge stand in the main hallway. Most people were walking straight past it and were gathering round the stands offering free Linux CDs. (I would really like to see at least one of the BSDs to add something like bproc - Linux has some amazing capabilities and those can and are used in many applications, but NetBSD has better I/O throughput and there are many cases where cluster applications are I/O-bound rather than CPU-bound or feature-bound.)
Microsoft's MPI implementation is, if I understand their materials correctly, based on MPICH (a BSD-licensed Open Source product) with some in-house fine tuning. MPICH is a good reference implementation but is not terribly fast and is getting to be long in the tooth. Far as I know, it doesn't have much in the way of fault tolerance in it, either. LAMPI and OpenMPI are built for speed (although I've found OpenMPI has room for substantial improvement) and have some fault tolerance support. So, they don't seem to be using an amazing architecture.
Last, but by no means least, Microsoft's freebies were limited to an Opteron-specific Windows 2003 Cluster Edition beta and a cookie. By comparison, many others had booklets on what their products did, papers on the theoretical work being done, working demos (the molecular modeler with forced feedback was amazing) and some highly knowledgeable geeks to answer detailed technical questions.
Microsoft may - someday - be an interesting player in the cluster market. Right now, though, they really don't seem to get what it is all about. I'm not trying to bash Microsoft here, they really don't have a product that is useful for the high-performance market, and seem to have the wrong libraries and interfaces for using the servers in a load-balancing, fail-over or distributed storage environment. This isn't to say the other vendors were perfect - I saw many areas that were horribly inefficient and poorly implemented - but rather that Microsoft would have done better to have come back from the show and re-thought what it was that they wanted the Cluster Edition to do.
I reversed the calculation to guess at how big the rock was that created the crater in Antarctica that was recently discovered, which is 300 miles across. I assumed that the asteroid had a fairly low density (porus rock). Assuming the object is travelling "slowly" (11 km/s), it would need to be 60 miles across to create a final crater of the necessary size. Even at 300 miles away (the edge of the crater), wind speeds would hit 8200 mph and the earth tremors would still be 11.3 on the Richter scale. A "typical" asteroid strike would be 17 km/s. To create the necessary crater, you'd be looking at a lump of rock 45 miles across. Most of the effects would be the same, except there would be a gigantic fireball. Again, at the crater's rim, you'd be looking at 8.53 x 10^10 joules/m^2 of energy for about 9110 seconds - enough to vaporize anything remotely close to the impact.
Assuming typical velocity, an iron asteroid would be a mere 22 miles across. The radiation would only be two-thirds that of the porus asteroid at the same speed.
If this was indeed the impact crater that triggered the initial phase of the Great Extinction, then the low density/high energy strike would produce vastly more heat and therefore affect the climate that much more.
...is that they're hoping that they can drag this on. IBM is making relatively little money off things like Linux - directly, at least - and it's unusual for defendents to be awarded costs in the US, so this is a drain on them that they can't recover. SCO must be hoping that IBM will eventually prefer to settle in order to cap the losses, as shareholders won't tolerate IBM paying out forever over something that won't earn so much as a dime.
If this is not their strategy, I can think of no possible reason for SCO to repeatedly delay proceedings, convolute issues or add further points of contention. It's not winning them any friends in the industry and I doubt it's helping their standing with the judge. They might be truly stupid, but it seems unlikely, and winning by default has a long and ignoble history.
After all, the S in SCO stands for Sauruman. (It can't be Sauron, as he won't allow his proper name to be written, according to Aragorn in "Two Towers".)
Ok, ok, I'll be serious for a second - it depends on what you define "advanced" as being relative to. If you're talking in the context of stone tools, then your lowest level is a lump of rock picked up from the ground, and your highest level is the finest the stone will fracture repeatably combined with the best possible shaping for the purpose of the tool. If you picture this as an X-Y graph, where (0,0) is the lump of rock and the ideal is (M,N), draw two diagonal lines where each starts on the Y axis, ends on the X axis and do not cross. Anything between (0,0) and that first diagonal is "primitive". Anything between the first diagonal and the second is "middling". Anything between the second diagonal and (M,N) is advanced.
The pygmies of Africa are relatively easy, using the conventional wisdom (ie: that mammals shrink on continents) - it's much harder to explain why there are so many taller people in Africa. Africa's a big place, with land ranging from deserts to thick jungle. Food is not a problem there, but avoiding becoming food is. The problem with a continent is that you have massive biodiversity, which means that there is a much higher likelihood of there being something capable of consuming, say, a human. Mammals have typically survived in such cases by shrinking, so as to be harder to see and to be better able to evade. It's not universal, but it's common enough to be considered a rule-of-thumb.
(Individual survival is also much more important, as there is a far lower probability for any given individual to survive and therefore far lower probability for a particular set of genes to carry on. In relatively isolated environments, such as islands, it is possible for communities of animals to develop a sense of cooperation, as there is no significant pressure on the individual EXCEPT from the community, and no real pressure on the community EXCEPT by reducing the size of the gene pool it can draw from. In an environment where scale is unlimited and individuals are under great stress, communities can afford to lose individuals - a large community won't be affected at all and small communities won't risk resources they don't have. Evolution will select for efficiency of the masses on islands, without regard to the efficiency of the individual, and will select for the efficiency of the individual in continents, without regard to the efficiency of groups.)
If he didn't say something else, it would hardly be hotly contested, would it?:) Besides, even when I was at University, Professors were forever disagreeing with me. Maybe it's something they're paid to do.:)
I would point out an obvious flaw in his logic, though - in order for something to be under development for 800,000 years (as opposed to merely being used for 800,000 years), you should see signs of progress. That's obvious, right? Well, typically what constitutes progress with a stone tool is the ability to produce a more controlled shape by reducing the minimum flake size you can produce. Stylizing tools (the "Clovis Point" being an example) can only happen once you're able to work at a fine enough level for such styles to be meaningful. Obviously, however, learning more effective styles is itself progress, but it is progress that can only occur after learning how to work the material to the necessary degree.
A "simple" stone tool implies no styling at all and only minimal progress in refining how to work the material. So, the question would be how many steps would it take for the improvement in workmanship to no longer be considered minimal? (Since a fine-grained reduction requires a fine-grained technique, initial steps would necessarily be coarse-grained. The rate of change of reduction in flake size would crudely approximate to a hyperbolic curve.) On the flip-side, how infrequent can improvements be made and still call things "progress" as opposed to "in use"?
Ultimately, improvements would seem to need to be frequent enough that, over the course of 800,000 years, it is simply no longer possible to call the tools "simple", except in relation to - say - later Iron Age or (in the case of this Professor) modern, industrialized, Western civilizations. Either that, or the progress WAS extremely limited, in which case the improvements could NOT be considered evidence of 800,000 years of progress. The professor can't have it both ways - it's either simple OR it's 800,000 years of R&D, it can't be both. The published data seems (to me) to support the conclusion that it was advanced enough to be beyond a culture of severely mentally impaired humans, no matter how long the society had.
(Early Stone Age cultures in Europe developed tools extremely slowly and most experts I've read on the subject usually state quite clearly that, for most of that time, technology was stagnant - to the point where it has been seriously debated as to whether early stone-age Europeans might have had a genetic defect leading to learning disabilities and/or severely crippled long-term memories.)
Near the site of the "hobbits", they have found fairly advanced stone tools and stone arrowheads. People so acutely affected by the suggested dwarfing disorder would have inhibited brain function and certainly could not have developed an advanced technology or operated it. This makes it somewhere between unlikely to impossible for all of the people on the island to have been mentally afflicted. This leaves only two options - either these remains are of extremely unusual people, and were in a community of more typical hominids, and it's pure chance that no remains of these typical hominids have been found, OR they genuinely were a miniaturized subcategory of hominid that were not impaired at all, so there is no contradiction involved with there being an advanced technology.
The debate has likely intensified even further with recent genetic studies of Neanderthals, using mtDNA extracted from the teeth. This is because the mtDNA shows vastly greater variation in early Neanderthal genetic makeup than had ever been expected. So much so that all prior studies are now considered grossly inadequate, as they only examined a hundred or so base pairs, considering the rest to be essentially identical. If genetic diversity in early hominids in general was as great as genetic diversity in early Neanderthals is believed to have been, then the probability of there having been a natural experiment in hobbits is considerably greater.
There is, however, one outstanding problem that has NOT been resolved. Dwarfism on islands is common with reptiles. Reptiles do NOT do islands well. However, mammals on islands tend towards giantism - Amblyrhiza Inundata (a giant rat the size of a grizzly) being an excellent example. Birds, although descended from reptiles, also seem to do well on islands - the Moa (a flightless bird that was 13 feet tall) and the Haast Eagle (the largest eagle that ever lived, with a wingspan of 14 feet), both from New Zealand, being good examples. This is because mammals scale well and therefore lose very little by being large, even when resources are scarce. Reptiles don't scale so well, so there is a loss of efficiency in being large. No big deal on a large enough land mass, but on an island, it's a major problem.
Humans, because they are potentially much better at cooperating, are capable of planning and storing, and are able to access a much wider range of foods over a much greater range of environments, should (based on knowledge of other island-based mammals) scale up on islands extremely well, and should only shrink where conflict is greatest, which would typically be a continent. It's hard to say if this is the case, as humans have always been amazingly mobile, but my gut feeling is that you'll find more very tall people on or around islands than you will in the middle of continents. This creates a problem for the hobbits, though. Mammals shrink when being able to run is a far greater survival trait than being able to gather more. On an island, there is very little to run from and almost nowhere to run to. There should, therefore, be no advantage to them being that small and therefore no reason for such a trait to be selected.
I think it likely that the hobbits are indeed a new branch of hominids, but without a good, solid explanation for why they would be small, the theory will never be acceptable to any evolutionary scientist worth a damn, no matter how much they want it to be true, simply because it runs counter to what we know about mammals on islands. Answering that one question will probably quell a lot of the more skeptical scientists, too. A mechanism that ties things together and presents a coherent picture is more acceptable than an extrapolation, no matter how many fossils it is from.
...there is no way to misread it as LARP. In the Spirit of Adventure system, the monsters used dye-soaked foam weapons. Made it easy to see the damage taken - usually even after three or four good long hand-washes later. (The sneaky monsters used extra dye, in the hopes that a really good hit would score double due to refs counting the splash-marks.)
Re:Take Two has never bothered much with quality.
on
Rockstar Plays it Safe
·
· Score: 1
Frontier was the sequel to Elite. It had some bugs but wasn't horrible. I forget which label that came out on. Frontier II (also known as Frontier: First Encounters) was the sequel to Frontier (making it the sesequel to Elite?:) and was released under the Take Two label, on behalf of David Braben and his development team. I believe there's more detail in the Elite newsgroup FAQ. Both Ian Bell and David Braben read (and post to) the newsgroup, BTW. (That's how I found out why some planets on the classic Elite have killer edible arts graduates.) You can probably find out more on the internals of Take Two by asking there... just don't be surprised if Braben's reply could be used as a replacement for nuclear power.
One of the more notorious scandals was Frontier II, which was (a) released before the coders were ready, (b) re-released with patch disks, (c) abandoned without warning rather than withdrawn, fixed and released properly. I have very little sympathy with Take Two over any of the messes they've got themselves into, although I do have sympathy with those gamers who have been subjected to their low-quality products.
Sure, they've released stuff that's been popular, but the only things that seem to be popular are the bits they've disabled.
Flamware might also be stuff made in the Flam fjord or a type of drum rudiment. Either of these would fit the discussion very nicely. Of course, the Real Ultimate Power version of flamware would be software where the license requires harsh criticism by means of sending flam-filled music bought in Flam.
It's not in the millions, it's only two hundred and fifty million, of whom a mere two million are active personnel. And how can it possibly result in identity theft when it only includes names, addresses, telephone numbers, social security numbers, claims information, data on spouses, housing information...
Actually, the best authentication mechanism for wireless at the moment is 802.1x. It is an encryption, authentication and access control mechanism that is per-user, and is often coupled with a standard secure mechanism such as Kerberos. If your wireless systems support 802.1x, that should be the system you use. No question about it.
If 802.1x is out of the question, then the next best solution is to use heavy encryption on the firewalls that connect the wireless access point to the rest of the network. Most firewalls will support IPSec. Running an IPSec tunnel between firewalls will eliminate any possibility of people breaking in OR of intercepts. On a wired network, I would consider IPSec to be the best security mechanism available. It suffers a little on wireless, because there is overhead involved, because the encryption is CPU-intensive on the firewalls and because it's sometimes tempermental on unreliable networks.
(Sun's SK/IP protocol was designed to overcome some of the heaviness of IPSec, but they abandoned it in a fit of pique when it didn't gain nearly the same levels of interest or enthusiasm.)
Two fairly early, but very potent and very powerful games. Revs has nothing that was unlockable, but the sheer challenge of it made the game good for a few months. Elite had two "secret missions", but those really didn't contribute much to the gameplay. The amazing difficulty of reaching that tech 1 anarchy when smuggling goods, or of being slammed by Thargons before being able to afford the military beam lasers... That game was good for a year or more, in that incarnation. There were two sequels, and an Open Source clone. (Several other open source clones were written, but vanished for one reason or another.)
To me, questions raised by these two games are not really answered by the modern practice of having almost everything secret in ways that cannot be logically figured out or logically encountered. Most of the time, people will get bored with looking for illogical stuff. If the game needs it, the player will lose interest. If the game doesn't need it, the player won't bother. Either that, or they wait for the information to be published on the Internet. Regardless, the secrets aren't adding anything and actually detract from things.
To me, logical sequencing and obviously in-character progressions are fine - even if they are secret. The secret nature isn't the important bit, though. They fit in that universe, they belong there, they make sense there, and they add to the feel of being there. It doesn't disrupt the flow - you aren't constantly switching between "playing" and "hitting things at random". I don't see any problem with such features and would expect them in many types of game. If these are a problem for the original article author, then we're definitely differing on what makes a game a game.
Another example from Way Back When - Level 9's text-based adventure games. These usually had one or more segments that were "massive", at least for the user. What they involved was a set of rooms that were re-used repeatedly, producing the illusion of a near-infinite space. They used some combination of colours and/or numbers to represent where in this virtual space you were. Once you realized that the space itself was a puzzle, it became fun to figure out how the system worked and, from that, infer where any secret exits would logically be. This fit with the dynamics of the game, so it just flowed naturally.
If I'm emphasising flow, it's because I believe this to be the characteristic of not only a good game, but also of a good strategy in a good game. To me, the number of secrets, the level of thought required, etc, are side issues. They alter the flavour of the game, but that's not what makes it good. Unlocking the logic of the game should be the chief puzzle, and once that has been conquered, the rest should largely follow.
The old ladders-style games had a very simple logic. Once you understood the timings, you didn't automatically master the game. That took practice, and each progressive level required greater precision. When this was done well, it kept the game interesting, because understanding was not the same as solving. It wasn't just ladders games that did this. The nastiest game in history has to be Firebird's Firetrack. It was a scrolling shoot-em-up that was FAST. After a game, you'd have to sit for several minutes until the walls stopped moving, and that was just the first level. Subsequent ones were faster and deadlier. Thinking wasn't an option and reflexes were too slow. If it actually had anything more to it than this, it could easily have thrashed anything else out at the time.
This is not to say that all older games were good. There was plenty of **** out there. Far TOO much. It's also not to say that no new games have this quality to them, but the percentage is far too low. There's no real complexity or challenge to many of them. Such games are difficult because they're obscure and not because of a titanic struggle for supremecy between coders and gamers.
You are in fact correct on both of these ways. Multicasting, as it exists on the backbone (though not on the ISPs!) is implemented by using a "group address" which operates by each router keeping a list of neighbors who are interested in that stream and then forwarding to only those neighbors, as you describe in method 2. When a computer joins one of these group addresses, its membership to that address is recorded by its local router. The router then joins that same group address on all the multicast routes it knows about. This continues upstream until a connection is made to other members of that group address OR the administrative distance is reached. Invalid routes are then pruned.
Actually, that describes one particular method of multicasting. There are others. Source-specific multicasting works by either specifying authorized or excluded host addresses, which obviously isn't efficient if the inclusion/exclusion data isn't passed up the tree, which means that routers end up with lists of endpoints that are valid, as per method 1.
You are also correct in that there are ways to get around non-multicast segments of the network. The two most common methods are tunneling (whereby you create a virtual private network that is "connected" to the closest multicast-aware router) and reflectors (which convert multicast streams into unicast ones and vice versa). Reflectors were very common when CU-SeeMe was the dominant videoconferencing tool on PCs and Macs - at the height of the CU-SeeMe craze, there were probably several tens of thousands of reflectors operating. Hey, that's not a bad total, when you consider that this was 7-12 years ago, broadband meant a 33K modem and the elite nerds were the ones with the color quickcams, not the b/w ones.
Although the MBone has been disbanded, you can still find some good information by looking it up. The older flood-and-prune protocol (DVMRP) has been replaced by something called the Protocol Independent Multicast protocol (PIM), of which there are several variants depending on the number of upstream and downstream connections you have. There are other protocols (MOSPF, Core Based Trees, and a few dozen specific to wireless networks where you have all kinds of added complexities!) but these don't appear to be in significant use. Again, worth reading up on, though.
Then you've various extensions, like Source-Specific Multicast (SSM), Scalable Reliable Multicast (SRM), and so on. SRM and other "reliable multicast" protocols are interesting because you can do things like HTTP or FTP over them, allowing you to transmit losslessly to a large number of clients at the same time. That would likely be very useful for massively distributed multiuser games, as it would be the only viable way to keep all the machines in sync once you start getting large numbers of users.
There are some excellent profiling products around (VTune, PAPI, DAKOTA, KOJAK, and those are just the ones I've used). Companies like VMWare probably use some form of profiling already - they'd be insane not to, as it's a great way to improve performance with little effort. Obviously, they'd be equal idiots if they published all of the stats churned out, but there are likely ways they could publish a set of general indexes or tables to show the overheads of running N OS' over M processors with P cores each, plus the cost of running some of the standard administrative functions. Because I'm talking about the low-level operation, rather than custom-made scenarios, the figures won't represent any given scenario exactly but can't be rigged by selecting a given example either. If other vendors then wanted to publish their own figures for the same matrix and functions, then people would have something to work with on comparisons.
Sure, you can probably plug the numbers into a suitably complicated equation, but it won't be linear and it won't be "obvious". The maxima won't be at the same place for different hypervisors, either. That's the point. If you use a single number benchmark, you can (almost) always find something product X does better than product Y. If you have the full behaviour of the system written out, vendors can't obscure things like that. It's good for the customer, as they can then see what product does the best with the specific characteristics they have in mind. It's also good for the vendor, because there's no pretense and no FUD (so the customers like you) and there's no denial (so the developers respect you).
Now, are ANY vendors going to do this? And I'm including Xen and VServers in this. Probably not. There are risks involved in being that transparent, plus costs. And even if the vendors all agreed it was a good idea, you think ANY of them would volunteer to go first?
This is not to diss VMWare. I respect them (as much as I respect any corporate entity) and this is just as true of the Open Source solutions. It's merely the practical reality that promoting a product through total education of the consumer is something neither party really wants. Customers want plausible denial if things don't work out, and vendors are not going to tell you to go to their competitors.
I don't see what the fuss is about, when it comes to planets, planetoids, etc. The problem is that astronomers have been using extremely an trivial value (diameter) to determine what to call non-stars, and use an equally trivial pair of values (spectral type and class) to determine star types. This seems to me to violate one of the core principles behind naming schemes (grouping in order to simplify understanding) and one of the fundamental tenants of science (keep things as simple as possible, but no simpler).
The Periodic Table of the Elements makes a lot of sense, because you can make a lot of predictions about the properties of an element based on where it is in the table. There are some oddities, sure, but by and large it is an extremely intuitive system. By comparison, knowing that a star is K or G tells you very little. You can make some inferences, by factoring in the abundances of the elements, the diameter of the star, the overall distribution of the electromagnetic radiation, etc, but if you're going to have to add in vast amounts of additional information to get anywhere, you might as well use that information in the name and have done with it.
For planets, asteroids, etc, it's much the same thing. By using too little information to determine the classification, you end up having to add vast amounts more information later on to produce subcategories, exceptions or new names entirely. That makes no sense to me whatsoever. Even a good naming system will need additions made to it, but it should be consistant with what is already there, and it should be easy to understand the relationships.
Since this is about planets, I'll use those as an illustration. Planets form around stars from the debris in the accretion disk, plus captured material from the stellar nursery in which the star formed, minus material "evaporated" from the system by the solar winds accelerating it, and minus material captured by other stars or gravitational sources. The process of condensing planets is slow, though apparently not as slow as once thought, which means that the material in the accretion disk will be sorted. In our own solar system, it seems to be that heavier elements are more common close to the sun and lighter ones are more common further away. (Mercury is unbelievably dense, for example, whereas Pluto seems to be little more than an iceball.)
However, because you need less energy to accelerate a lower mass, and because elemental hydrogen only forms a solid under extreme pressures, these will ALL have abundances of elements that are skewed (possibly by a lot, for inner planets, as the solar winds are much stronger) from the ratios observed on much larger scales (say, in the galaxy or the observable universe). Stars, on the other hand, are mostly composed of the extremely light elements and fit the expected abundances very nicely. As the gravitational field is reduced, the skew should increase, as it would require that much less energy for something to be ripped away, if it's free. (Obviously, hydrogen that has reacted with oxygen to form water is going to require much more energy than elemental hydrogen alone.) So, the composition tells us a lot about where something forms, how quickly it accumulated mass and how long it took. It would seem obvious, then, that composition should bear a major role in deciding what to call something.
The other "obvious" one would be structure. The "asteroid" recently observed to be 45% empty space (sand is 25%) would probably merit a new classification. Most asteroids probably have multiple "centers" around which they have congealed/collided. Certainly, the two comets that have broken up have had multiple centers, not a single rocky core. By comparison, the gas giants have a single center (duh!), as does the Earth and Venus, probably Mars as well, not sure if there's enough data on the others. But even with that, we can clearly see a logical distinction (as opposed to an arbitrary one) that can clearly distinguish between two very
Slashdot Mirror
The other thing to consider is that it isn't easy to determine ability in a subject. The best mechanism I can think of is to perform per-subject tests at regularly-spaced intervals, where the test has no upper limit on scores, weighted by the stream they are currently in. The closer to the expected weighted value for that student, the better the fit between stream and student. A score above expectations means the student is quite capable of learning faster. A score below expectations means either a lack of motivation or a lack of ability at that pace. That cannot easily be resolved by tests and would be better resolved by working with the student. A lack of motivation through boredom should result in a faster stream. A lack of ability should mean a slower stream. Failure to communicate well with the teacher should mean going to the nearest stream either side in which communication is improved.
Because this is a massively multi-variable problem, an ideal fit is unlikely for all students. The best anyone can hope for is a superior fit for the majority of students, with no worse than "standard" methods for all others.
...they did have to search for it.
I am not interested in qualitative studies for the most part and I have little regard for statistical studies. These are too easily rigged to suit people's preconceptions. Empirical research - where possible - where a definite pathway A-to-B-to-C can be demonstrated is the ideal. In education, empirical research is extremely difficult and it's only recently that technology such as fMRI even existed. Certainly, schools aren't equipt with such technology to routinely scan the kids' brains to determine how the school culture is influencing things. Virtually all other research is culturally biased and heavily statistical with no statement on the level of confidence used in the statistical tests, or indeed what statistical method was used, and how it was ensured that a genuinely representative random sample was obtained.
The understanding of the importance of language comes from a mix of research - some involving fMRI (which lends credibility to it), but most involving qualitative assessments of education and the impact on society from schools in Europe over the past 2,500 years. The amount of data is fairly impressive and there is therefore some credibility beyond the "hard science" to the argument that multi-cultural and linguistically sophisticated education is far superior to insular mono-cultures such as those found in many parts of the US.
I'll finish up with costs. I hold to the belief that skilled work generally produces more than it costs, whereas unskilled work is merely necessary to get any work done at all. If you eliminate unskilled labour as much as possible, transferring it to machines or whatever, and raise the educational standards across the board, the net value of the work done will rise. Since we pay taxes as a percent of our income, and corporations pay taxes according to their earnings, etc, the net value of taxes must also rise, if the populace is adequately educated. Now, the law of diminishing returns does come into play here. You can't improve education forever and expect the wealth generated to go to infinity. That won't happen. What will happen is that it will tend to some upper limit. There is therefore some upper bound where further investment will have no significant benefit. However, investment BELOW that point will generate a substandard return and investing further will reap enough of a return to be profitable all round.
(It's not a simple relationship. Greater education produces not only greater skills but better research, which means that the cost for R&D will fall, with respect to the value of the products developed. This will directly benefit the companies, but because you now have more money circulating, you also have more money collected in taxes. R&D is much more random, however, and therefore it is much harder to predict the impact of higher-quality work. This is something you can really only try and see.)
Total investment in education is as much a myth as the total employmen
Based on my own volunteer work in school programs, I would say that class sizes should rarely be above 15-20 in total, and should have 1 teacher/assistant competent in the subject for every 5-7 students. I also think kids should be streamed per subject, with some flexibility for when certain groups of kids happen to work well together. (No, that does not mean cribbing the notes.)
The problem with the existing system is that it is geared around people learning as and when the teacher gets round to it, rather than pushing people as far and as fast as they are able. It is no wonder that kids use drugs, but my guess is that its more to zone out the inadequacies of the educational system as it is to improve learning. You can't accelerate much beyond the speed the material is taught.
Based on research that has been caried out, I think that I'd extend this basic concept by throwing in a second or even a third language, as it appears that the complexity of language is such that learning new languages young boosts the growth of neural connections and seems to improve the capacity to learn. Languages, therefore, may provide a safe alternative to these drugs in that they'll boost intelligence and have no risk of later side-effects.
Difficulty, therefore, is NOT a significant factor in all of this. Ok, what about expense? Well, you're right that Linux is free. So is OpenMOSIX, OpenMPI (and many other MPI implementations), PVM (another messaging library), Lustre (a very high-performance network file system), many scientific and mathematical applications for clusters, etc. There are clustering patches for PoVRay, and it's always possible to write a script to have multiple machines render parts of images anyway. I'm sure there are other applications out there that I'm not thinking of right now, and it's only a matter of time before more "mundane" applications can take advantage of clustered environments. They already do, on Plan 9, to some degree. Oh, Plan 9 is also free.
Cost would appear not to be a major problem either, then. Optimizing is the only thing that is in any way difficult, and a GUI system that doesn't let you get to the really fine detail won't help there. More time, effort and money is spent on optimizing than on anything else, and I simply can't see any possible way that an OS that is designed for ease-of-use by hiding the intricacies can in any way help in that.
Microsoft's MPI implementation is, if I understand their materials correctly, based on MPICH (a BSD-licensed Open Source product) with some in-house fine tuning. MPICH is a good reference implementation but is not terribly fast and is getting to be long in the tooth. Far as I know, it doesn't have much in the way of fault tolerance in it, either. LAMPI and OpenMPI are built for speed (although I've found OpenMPI has room for substantial improvement) and have some fault tolerance support. So, they don't seem to be using an amazing architecture.
Last, but by no means least, Microsoft's freebies were limited to an Opteron-specific Windows 2003 Cluster Edition beta and a cookie. By comparison, many others had booklets on what their products did, papers on the theoretical work being done, working demos (the molecular modeler with forced feedback was amazing) and some highly knowledgeable geeks to answer detailed technical questions.
Microsoft may - someday - be an interesting player in the cluster market. Right now, though, they really don't seem to get what it is all about. I'm not trying to bash Microsoft here, they really don't have a product that is useful for the high-performance market, and seem to have the wrong libraries and interfaces for using the servers in a load-balancing, fail-over or distributed storage environment. This isn't to say the other vendors were perfect - I saw many areas that were horribly inefficient and poorly implemented - but rather that Microsoft would have done better to have come back from the show and re-thought what it was that they wanted the Cluster Edition to do.
Assuming typical velocity, an iron asteroid would be a mere 22 miles across. The radiation would only be two-thirds that of the porus asteroid at the same speed.
If this was indeed the impact crater that triggered the initial phase of the Great Extinction, then the low density/high energy strike would produce vastly more heat and therefore affect the climate that much more.
But... but... if it hits the market, it's crashed and is no longer flying. Surely you want a car that DOESN'T hit the market?
Hmmm. Ok, those are crap, so I guess Wireshark isn't too awful.
If this is not their strategy, I can think of no possible reason for SCO to repeatedly delay proceedings, convolute issues or add further points of contention. It's not winning them any friends in the industry and I doubt it's helping their standing with the judge. They might be truly stupid, but it seems unlikely, and winning by default has a long and ignoble history.
After all, the S in SCO stands for Sauruman. (It can't be Sauron, as he won't allow his proper name to be written, according to Aragorn in "Two Towers".)
Ok, ok, I'll be serious for a second - it depends on what you define "advanced" as being relative to. If you're talking in the context of stone tools, then your lowest level is a lump of rock picked up from the ground, and your highest level is the finest the stone will fracture repeatably combined with the best possible shaping for the purpose of the tool. If you picture this as an X-Y graph, where (0,0) is the lump of rock and the ideal is (M,N), draw two diagonal lines where each starts on the Y axis, ends on the X axis and do not cross. Anything between (0,0) and that first diagonal is "primitive". Anything between the first diagonal and the second is "middling". Anything between the second diagonal and (M,N) is advanced.
(Individual survival is also much more important, as there is a far lower probability for any given individual to survive and therefore far lower probability for a particular set of genes to carry on. In relatively isolated environments, such as islands, it is possible for communities of animals to develop a sense of cooperation, as there is no significant pressure on the individual EXCEPT from the community, and no real pressure on the community EXCEPT by reducing the size of the gene pool it can draw from. In an environment where scale is unlimited and individuals are under great stress, communities can afford to lose individuals - a large community won't be affected at all and small communities won't risk resources they don't have. Evolution will select for efficiency of the masses on islands, without regard to the efficiency of the individual, and will select for the efficiency of the individual in continents, without regard to the efficiency of groups.)
I would point out an obvious flaw in his logic, though - in order for something to be under development for 800,000 years (as opposed to merely being used for 800,000 years), you should see signs of progress. That's obvious, right? Well, typically what constitutes progress with a stone tool is the ability to produce a more controlled shape by reducing the minimum flake size you can produce. Stylizing tools (the "Clovis Point" being an example) can only happen once you're able to work at a fine enough level for such styles to be meaningful. Obviously, however, learning more effective styles is itself progress, but it is progress that can only occur after learning how to work the material to the necessary degree.
A "simple" stone tool implies no styling at all and only minimal progress in refining how to work the material. So, the question would be how many steps would it take for the improvement in workmanship to no longer be considered minimal? (Since a fine-grained reduction requires a fine-grained technique, initial steps would necessarily be coarse-grained. The rate of change of reduction in flake size would crudely approximate to a hyperbolic curve.) On the flip-side, how infrequent can improvements be made and still call things "progress" as opposed to "in use"?
Ultimately, improvements would seem to need to be frequent enough that, over the course of 800,000 years, it is simply no longer possible to call the tools "simple", except in relation to - say - later Iron Age or (in the case of this Professor) modern, industrialized, Western civilizations. Either that, or the progress WAS extremely limited, in which case the improvements could NOT be considered evidence of 800,000 years of progress. The professor can't have it both ways - it's either simple OR it's 800,000 years of R&D, it can't be both. The published data seems (to me) to support the conclusion that it was advanced enough to be beyond a culture of severely mentally impaired humans, no matter how long the society had.
(Early Stone Age cultures in Europe developed tools extremely slowly and most experts I've read on the subject usually state quite clearly that, for most of that time, technology was stagnant - to the point where it has been seriously debated as to whether early stone-age Europeans might have had a genetic defect leading to learning disabilities and/or severely crippled long-term memories.)
The debate has likely intensified even further with recent genetic studies of Neanderthals, using mtDNA extracted from the teeth. This is because the mtDNA shows vastly greater variation in early Neanderthal genetic makeup than had ever been expected. So much so that all prior studies are now considered grossly inadequate, as they only examined a hundred or so base pairs, considering the rest to be essentially identical. If genetic diversity in early hominids in general was as great as genetic diversity in early Neanderthals is believed to have been, then the probability of there having been a natural experiment in hobbits is considerably greater.
There is, however, one outstanding problem that has NOT been resolved. Dwarfism on islands is common with reptiles. Reptiles do NOT do islands well. However, mammals on islands tend towards giantism - Amblyrhiza Inundata (a giant rat the size of a grizzly) being an excellent example. Birds, although descended from reptiles, also seem to do well on islands - the Moa (a flightless bird that was 13 feet tall) and the Haast Eagle (the largest eagle that ever lived, with a wingspan of 14 feet), both from New Zealand, being good examples. This is because mammals scale well and therefore lose very little by being large, even when resources are scarce. Reptiles don't scale so well, so there is a loss of efficiency in being large. No big deal on a large enough land mass, but on an island, it's a major problem.
Humans, because they are potentially much better at cooperating, are capable of planning and storing, and are able to access a much wider range of foods over a much greater range of environments, should (based on knowledge of other island-based mammals) scale up on islands extremely well, and should only shrink where conflict is greatest, which would typically be a continent. It's hard to say if this is the case, as humans have always been amazingly mobile, but my gut feeling is that you'll find more very tall people on or around islands than you will in the middle of continents. This creates a problem for the hobbits, though. Mammals shrink when being able to run is a far greater survival trait than being able to gather more. On an island, there is very little to run from and almost nowhere to run to. There should, therefore, be no advantage to them being that small and therefore no reason for such a trait to be selected.
I think it likely that the hobbits are indeed a new branch of hominids, but without a good, solid explanation for why they would be small, the theory will never be acceptable to any evolutionary scientist worth a damn, no matter how much they want it to be true, simply because it runs counter to what we know about mammals on islands. Answering that one question will probably quell a lot of the more skeptical scientists, too. A mechanism that ties things together and presents a coherent picture is more acceptable than an extrapolation, no matter how many fossils it is from.
...there is no way to misread it as LARP. In the Spirit of Adventure system, the monsters used dye-soaked foam weapons. Made it easy to see the damage taken - usually even after three or four good long hand-washes later. (The sneaky monsters used extra dye, in the hopes that a really good hit would score double due to refs counting the splash-marks.)
Frontier was the sequel to Elite. It had some bugs but wasn't horrible. I forget which label that came out on. Frontier II (also known as Frontier: First Encounters) was the sequel to Frontier (making it the sesequel to Elite? :) and was released under the Take Two label, on behalf of David Braben and his development team. I believe there's more detail in the Elite newsgroup FAQ. Both Ian Bell and David Braben read (and post to) the newsgroup, BTW. (That's how I found out why some planets on the classic Elite have killer edible arts graduates.) You can probably find out more on the internals of Take Two by asking there... just don't be surprised if Braben's reply could be used as a replacement for nuclear power.
Sure, they've released stuff that's been popular, but the only things that seem to be popular are the bits they've disabled.
Flamware might also be stuff made in the Flam fjord or a type of drum rudiment. Either of these would fit the discussion very nicely. Of course, the Real Ultimate Power version of flamware would be software where the license requires harsh criticism by means of sending flam-filled music bought in Flam.
Oh.
If 802.1x is out of the question, then the next best solution is to use heavy encryption on the firewalls that connect the wireless access point to the rest of the network. Most firewalls will support IPSec. Running an IPSec tunnel between firewalls will eliminate any possibility of people breaking in OR of intercepts. On a wired network, I would consider IPSec to be the best security mechanism available. It suffers a little on wireless, because there is overhead involved, because the encryption is CPU-intensive on the firewalls and because it's sometimes tempermental on unreliable networks.
(Sun's SK/IP protocol was designed to overcome some of the heaviness of IPSec, but they abandoned it in a fit of pique when it didn't gain nearly the same levels of interest or enthusiasm.)
To me, questions raised by these two games are not really answered by the modern practice of having almost everything secret in ways that cannot be logically figured out or logically encountered. Most of the time, people will get bored with looking for illogical stuff. If the game needs it, the player will lose interest. If the game doesn't need it, the player won't bother. Either that, or they wait for the information to be published on the Internet. Regardless, the secrets aren't adding anything and actually detract from things.
To me, logical sequencing and obviously in-character progressions are fine - even if they are secret. The secret nature isn't the important bit, though. They fit in that universe, they belong there, they make sense there, and they add to the feel of being there. It doesn't disrupt the flow - you aren't constantly switching between "playing" and "hitting things at random". I don't see any problem with such features and would expect them in many types of game. If these are a problem for the original article author, then we're definitely differing on what makes a game a game.
Another example from Way Back When - Level 9's text-based adventure games. These usually had one or more segments that were "massive", at least for the user. What they involved was a set of rooms that were re-used repeatedly, producing the illusion of a near-infinite space. They used some combination of colours and/or numbers to represent where in this virtual space you were. Once you realized that the space itself was a puzzle, it became fun to figure out how the system worked and, from that, infer where any secret exits would logically be. This fit with the dynamics of the game, so it just flowed naturally.
If I'm emphasising flow, it's because I believe this to be the characteristic of not only a good game, but also of a good strategy in a good game. To me, the number of secrets, the level of thought required, etc, are side issues. They alter the flavour of the game, but that's not what makes it good. Unlocking the logic of the game should be the chief puzzle, and once that has been conquered, the rest should largely follow.
The old ladders-style games had a very simple logic. Once you understood the timings, you didn't automatically master the game. That took practice, and each progressive level required greater precision. When this was done well, it kept the game interesting, because understanding was not the same as solving. It wasn't just ladders games that did this. The nastiest game in history has to be Firebird's Firetrack. It was a scrolling shoot-em-up that was FAST. After a game, you'd have to sit for several minutes until the walls stopped moving, and that was just the first level. Subsequent ones were faster and deadlier. Thinking wasn't an option and reflexes were too slow. If it actually had anything more to it than this, it could easily have thrashed anything else out at the time.
This is not to say that all older games were good. There was plenty of **** out there. Far TOO much. It's also not to say that no new games have this quality to them, but the percentage is far too low. There's no real complexity or challenge to many of them. Such games are difficult because they're obscure and not because of a titanic struggle for supremecy between coders and gamers.
I believe it would be easy enough to test t
Actually, that describes one particular method of multicasting. There are others. Source-specific multicasting works by either specifying authorized or excluded host addresses, which obviously isn't efficient if the inclusion/exclusion data isn't passed up the tree, which means that routers end up with lists of endpoints that are valid, as per method 1.
You are also correct in that there are ways to get around non-multicast segments of the network. The two most common methods are tunneling (whereby you create a virtual private network that is "connected" to the closest multicast-aware router) and reflectors (which convert multicast streams into unicast ones and vice versa). Reflectors were very common when CU-SeeMe was the dominant videoconferencing tool on PCs and Macs - at the height of the CU-SeeMe craze, there were probably several tens of thousands of reflectors operating. Hey, that's not a bad total, when you consider that this was 7-12 years ago, broadband meant a 33K modem and the elite nerds were the ones with the color quickcams, not the b/w ones.
Although the MBone has been disbanded, you can still find some good information by looking it up. The older flood-and-prune protocol (DVMRP) has been replaced by something called the Protocol Independent Multicast protocol (PIM), of which there are several variants depending on the number of upstream and downstream connections you have. There are other protocols (MOSPF, Core Based Trees, and a few dozen specific to wireless networks where you have all kinds of added complexities!) but these don't appear to be in significant use. Again, worth reading up on, though.
Then you've various extensions, like Source-Specific Multicast (SSM), Scalable Reliable Multicast (SRM), and so on. SRM and other "reliable multicast" protocols are interesting because you can do things like HTTP or FTP over them, allowing you to transmit losslessly to a large number of clients at the same time. That would likely be very useful for massively distributed multiuser games, as it would be the only viable way to keep all the machines in sync once you start getting large numbers of users.
Sure, you can probably plug the numbers into a suitably complicated equation, but it won't be linear and it won't be "obvious". The maxima won't be at the same place for different hypervisors, either. That's the point. If you use a single number benchmark, you can (almost) always find something product X does better than product Y. If you have the full behaviour of the system written out, vendors can't obscure things like that. It's good for the customer, as they can then see what product does the best with the specific characteristics they have in mind. It's also good for the vendor, because there's no pretense and no FUD (so the customers like you) and there's no denial (so the developers respect you).
Now, are ANY vendors going to do this? And I'm including Xen and VServers in this. Probably not. There are risks involved in being that transparent, plus costs. And even if the vendors all agreed it was a good idea, you think ANY of them would volunteer to go first?
This is not to diss VMWare. I respect them (as much as I respect any corporate entity) and this is just as true of the Open Source solutions. It's merely the practical reality that promoting a product through total education of the consumer is something neither party really wants. Customers want plausible denial if things don't work out, and vendors are not going to tell you to go to their competitors.
The Periodic Table of the Elements makes a lot of sense, because you can make a lot of predictions about the properties of an element based on where it is in the table. There are some oddities, sure, but by and large it is an extremely intuitive system. By comparison, knowing that a star is K or G tells you very little. You can make some inferences, by factoring in the abundances of the elements, the diameter of the star, the overall distribution of the electromagnetic radiation, etc, but if you're going to have to add in vast amounts of additional information to get anywhere, you might as well use that information in the name and have done with it.
For planets, asteroids, etc, it's much the same thing. By using too little information to determine the classification, you end up having to add vast amounts more information later on to produce subcategories, exceptions or new names entirely. That makes no sense to me whatsoever. Even a good naming system will need additions made to it, but it should be consistant with what is already there, and it should be easy to understand the relationships.
Since this is about planets, I'll use those as an illustration. Planets form around stars from the debris in the accretion disk, plus captured material from the stellar nursery in which the star formed, minus material "evaporated" from the system by the solar winds accelerating it, and minus material captured by other stars or gravitational sources. The process of condensing planets is slow, though apparently not as slow as once thought, which means that the material in the accretion disk will be sorted. In our own solar system, it seems to be that heavier elements are more common close to the sun and lighter ones are more common further away. (Mercury is unbelievably dense, for example, whereas Pluto seems to be little more than an iceball.)
However, because you need less energy to accelerate a lower mass, and because elemental hydrogen only forms a solid under extreme pressures, these will ALL have abundances of elements that are skewed (possibly by a lot, for inner planets, as the solar winds are much stronger) from the ratios observed on much larger scales (say, in the galaxy or the observable universe). Stars, on the other hand, are mostly composed of the extremely light elements and fit the expected abundances very nicely. As the gravitational field is reduced, the skew should increase, as it would require that much less energy for something to be ripped away, if it's free. (Obviously, hydrogen that has reacted with oxygen to form water is going to require much more energy than elemental hydrogen alone.) So, the composition tells us a lot about where something forms, how quickly it accumulated mass and how long it took. It would seem obvious, then, that composition should bear a major role in deciding what to call something.
The other "obvious" one would be structure. The "asteroid" recently observed to be 45% empty space (sand is 25%) would probably merit a new classification. Most asteroids probably have multiple "centers" around which they have congealed/collided. Certainly, the two comets that have broken up have had multiple centers, not a single rocky core. By comparison, the gas giants have a single center (duh!), as does the Earth and Venus, probably Mars as well, not sure if there's enough data on the others. But even with that, we can clearly see a logical distinction (as opposed to an arbitrary one) that can clearly distinguish between two very