To all those bitching about how a $50 discount won't convert any Windows users: you're missing the point.
Dell is in shit, in a very competitive, saturated market. Common business sense says the best thing they can do is to be the first to into a new market.
Headline: the new market is not one of potential converts from Windows - not yet, anyhow.
Dell's hoped-for market consists of :
* long-term linux users with Real Jobs who need a new PC but don't have the time or just can't be arsed to build, install, and customise Yet Another Linux Box.
* The type who sometimes, against their best wishes, might feel themselves lured back to Windows because they feel it's just *easier*.
* The type who would quite like official support on a preinstalled linux box, instead of banging their heads against the ITSupport wall when the Windows box they just bought, wiped, reinstalled & voided the warranty for has a hardware problem.
These types are likely to discriminate in the long term against a company that quoted them *more* for a preinstalled linux box. For these types, $50 is not an outright incentive, but it "is" a nice goodwill gesture. Converting Windows users comes later.
Ladies and gentlemen, Dell's intended market is *us*. And no-one's bothered with us before.
As the man said elsewhere : Vote. Wallet. Now. I know I will as soon as the UK gets the same offer.
> Since this seems like such a new concept (please correct me if I'm wrong; I > don't follow plane technology too much), it would just seem prudent to try > bending the wings until they break...
There's lots of reasons not to test a CFRP wing to failure.
1/ Wing test rigs are designed to break standard metallic wings. Rigs of this scale are *not* designed for wings that deform this much - some fancy rig redesign would be necessary to keep the loading plane normal on a wing that can double back on itself!
2/ Aerostructural failure is already defined by the limit load & ultimate load criteria. Basically, limit load is defined by certification authority book cases, and is the maximum load the aircraft would be expected to see in service, plus some. There can be no detrimental deformation at limit load. Above limit load, separation of structure must not occur, although repair may be required. Ultimate load is the criterion defined as 150% of limit load (for civil aircraft - it's different for military); if an a/c is ever subjected to ultimate load then it should be retired. Therefore there's no "prudence" in testing a wing (much) above ultimate load; the only purpose it would serve is to check if your calcs are overly conservative and/or your manufacturing & assembly is up to scratch.
3/ A standard a/c design (such as every civil airliner currently on Earth) would fall out of the sky for aerodynamic reasons long, long before its wingtips touched. In fact, it's possible that much of the wing would have to be overstrength just to maintain the stiffness required for aerodynamic purposes.
4/ As pointed out elsewhere in this thread, in metallic structure energy is dissipated as plastic deformation (& heat); carbon fibre is almost purely elastic, hence *none* of the energy is lost until fracture. I wouldn't want to be in the same room as that large a chunk of CFRP containing enough energy to bend a wing through 180 degrees (frankly I wouldn't even want to be on the clean-up crew afterwards - all that dust)! This is the same reason that metallic pressure vessels in chemical works are tested hydraulically, not pneumatically (gas = compressive = 1 big spring).
5/ The criterion is that any single component has how can they make accurate judgments and calculations without knowing exactly > how much stress the wings can take before snapping?
Point 3/ explains why a specific wing's breaking limit is irrelevant.
However, you kind of have a point in terms of correlation of calculation methods with known failure, especially because composite technologies are so new, unknown, and CFRP structure may be (probably is) overdesigned.
The reality of overtesting a composite wing is much more complex, though.
Yes, continuing a static test 'til failure would provide some useful feedback on analysis techniques. Not as much as you'd think, though:
a/ Even in metallic structure there's a lot of scatter associated with material properties, assembly, etc. There are several detail areas in aircraft design where the calcs are known to be conservative: although removing conservatism can increase a/c efficiency, it is ultimately more important to design robustly if a detail is very sensitive to small changes in local load.
b/ Composite strength is affected by a huge number of issues that don't affect metals. Manufacturing & assembly errors can cause problems (in effect you create the material at the same time as you make your structure). Moisture ingress into the matrix can affect material strength considerably and also add weight; prolonged exposure to fuel will also degrade the material. 2D CFRP laminates are very intolerant to contact and impact, so birdstrike or damage from runway debris will also weaken the material. To summarise, there are a number of new factors, all have to be accounted for, which leads to a wing that is several times stronger than it needs to be, simply to ensure that it's *still* strong enough on its
> Additionally, it was determined that the concrete foundations were binding > CO2 as they cured (concrete cures for years and years), causing still more > problems"
No. Dangerously high levels of CO2 terminated the biosphere2 project prematurely.
The cause was a naive assumption during dome construction that there was no upper limit to the degree of soil fertility in a sealed environment: when preparing the soil, it was decided to boost the vegetation (hence O2 levels) by digging in lots of extra organic matter - I think they tripled the originally-quoted quantity. The slight increase in O2 output was more than offset by skyrocketing CO2 levels from the decaying organics. The inhabitants suffered severe CO2 poisoning, could barely continue functioning, and eventually had to crack the seal.
Subsequent investigations showed a large mass of CO2 was unaccounted for in the dome's atmosphere. That's when someone realised that the strange chemical changes that had been observed in the concrete structure had actually been sequestering the CO2; the concrete was slowly turning to something like limestone.
Too much organic matter doomed the project to failure. The concrete structure actually *extended* its life considerably. As far as I know, no studies were conducted to examine the long-term effects on the viability of the structure.
I can't agree with the implied criticism of "MirrorMask". I simply loved it, so did my wife.
Sure it lacked mass appeal; it was very much an "art-house" feel, and the cast of only 7 or so people ensured that the film felt more like a screenplay (not necessarily a bad thing in my opinion - check out "Dogville").
But the plot was inventive and quintessentially Gaiman, and the acting was top notch. Costumes and sets were simple but well thought out, and seeing Dave McKean's usual style of artwork transformed from the static page to cgi was absolutely jaw-dropping; it felt like 3D, somehow.
Weirder, both the wife and I had really powerful, lucid dreams that night. How many blockbusters can you say *that* about?
"Mr. Shuttleworth is to be applauded for not bending to Microsoft." "It's pleasing to hear that he will not be directly working with Microsoft" "Dear Ballmer. NUTS. Signed, Mark Shuttleworth"
Ok, so Shuttleworth hasn't taken MS's shilling on the patent issue: but he hasn't exactly given a "Nuts" reply either, and has most certainly not discounted directly working with MS, quite the opposite.
From Shuttleworth's blog:
1/ "We have declined to discuss any agreement with Microsoft under the threat of unspecified patent infringements,"
2/ "Allegations of 'infringement of unspecified patents' carry no weight whatsoever. We don't think they have any legal merit, and they are no incentive for us to work with Microsoft on any of the wonderful things we could do together. A promise by Microsoft not to sue for infringement of unspecified patents has no value at all and is not worth paying for."
3/ "I have no objections to working with Microsoft in ways that further the cause of free software," "and I don't rule out any collaboration with them, in the event that they adopt a position of constructive engagement with the free software community."
I read this as :
1/ "We *have* been approached Microsoft, who insisted on the patent clause as part of the deal" .
2/ "Just in case MS thinks that our refusal is just us stalling for a higher price, let me make it clear: pull the other one, guys, it's got bells on it. *Any* price you offer will not reimburse the value lost to us if we publically state that MS has IP in Linux OS".
3/ "If MS really *are* for real on the question of interoperability, then we're still listening.
Shuttleworth is refusing to capitulate on the patent issue, because if he does he signs away the value of his company. He calls MS' bluff by implying that Canonical would insist on opening the relevant interoperability source, but says it in a way that if MS are for real, then no-one's lost face and negotiations can proceed.
My estimation of Mr Shuttleworth's business acumen just shot up.
> > Personally I detest the proprietary shit its peddling, but Ubuntu's already proven there's > > a demand for that. > Ubuntu went so far as to create a "restricted driver manager" that tells you when you're > using binary drivers, why you shouldn't, and what you can do about it.
You appear to be confused.
I didn't criticise Ubuntu. I actually quite *like* Ubuntu's pragmatic take on proprietary drivers:
* They're in the distro because they work, and there's currently no other option other than breaking functionality. * This helps win Windows users over, who want their systems to Just Work (tm). Linux user base grows. * New users are then educated by the restricted driver manager. * Completely free version of Ubuntu is in the pipeline.
I think this is a smart way forward, and Ubuntu will be my 2nd choice of distro if Fedora ever goes seriously downhill or bites the dust.
I *still* hate the proprietary shit. Have done for years, ever since I updated my kernel, yum couldn't find a matching NVidia 3D binary driver, and my next reboot left me at a text console. Can't wait 'til Nouveau is ready for the mainstream.
> If you call this "peddling proprietary shit", then I don't think you understand more than > one of those words.
I understand just fine: you misread my post, and thought I was criticising what's obviously favourite distro (Ubuntu) when I was actually criticising CNR. So you then posted a heated personal reply.
Like Ubuntu: *hate* proprietary drivers & codecs. Is that clear enough?
> Who do you reakon will be the next distro to enter into a Microsoft deal. So far its been > all the ones you'd expect - I suspect we won't be hearing much more of Xandros and Linspire, > although I think Novell could still come out unscathed- My guess is Mandriva - although it > saddens me to say it - they are having financial troubles etc.
I suspect you're right. But then again the strong French socialist streak may win through.
> Mark Shuttleworth said he wouldn't go into any kind of deal like this and I think I believe > him.
I definitely do - it would be a travesty of all the reasons he says he got into the business, and he's not exactly strapped for beer money.
> Most of all we've got to hope that Red Hat doesn't,
If they do then it will be a pyrrhic victory for MS; I suspect the other distros would shortly be welcoming a lot of new kernel developers into their ranks. Or we might see the start of another rpm-based Linux distro. What's the value of a Linux company without its engineers/developers?
Besides, MS don't *need* to take out Redhat. By turning Novell they already threw a spanner in the works for the Linux Standards Base; if Linux is already fragmenting why draw any additional antitrust heat?
> as the largest commercial Linux company it would be disastrous for any possible defence we > have against possible patent issues etc
Not really.
Redhat has already released virtually all their code - CentOS is a good showcase for what's left when you strip out everything proprietary.
As to patent issues, who at Redhat do you think chases those? The business guys? Really?
Clue: ever wondered why Alan Cox (FSF Award winner 2003, anti-s/w patent activist) isn't as code-productive as he was a few years ago?
In a market where there is an infinite supply of the raw product (e.g. baked beans), you have to compete on "value added", which often manifests a perceived quality brand. In the software world the engineers & developers are what add value.
Any linux distro that loses its developers is dead in the water - check out SCO.
Put it another way. You're worrying too much about the corporates; keep your eye on their staff instead.
>...continues to hang itself. At least we still have Debian. Even though its derivatives will > probably all sell out.
A little perspective here, please.
* Novell sold out because, despite their purchases of Ximian & SuSE, they never really "got" Linux; they were just trying to shore up a rapidly dying Netware product while continuing business along the same paradigms that killed it. Witness the resulting exodus of several core SUSE developers, further reducing the company's understanding of Linux; frankly I've seen Novell Linux brands as almost defunct for some years now. (go on, flame away...)
* Xandros sold out because their market share and community was miniscule. They sought to steal Windows market share, but (unsurprisingly) didn't have the resource to tackle Redmond. Xandros are already defunct and starting to smell; they just don't know it yet. (go on, flame some more...)
* Linspire haven't really recovered since having their teeth pulled, and they really don't "get" the security issue. The whole distro is very much Kevin Carmony's baby, and seems to be very fluid while it tries to find a profitable niche. Ubuntu's just broken into the territory it was trying to win (i.e. preinstalled mainstream linux), so I think the distro will die soon. Strangely, though, I don't think that Linspire has sold out, exactly, it's following in its father's footsteps; it understands business, not OS, and is evolving into a kind of "software accessibility enabler". Personally I detest the proprietary shit its peddling, but Ubuntu's already proven there's a demand for that.
So MS has munched on the low-hanging fruit. Sad, but not unexpected; the old & weak are always the first to go in war & business. What remains is :
* Several hundred non-commercial distros, top of the list is Debian, the epitome of idealism. * Ubuntu - very smart, idealistic, breaking into the mainstream. * Redhat - very smart, idealistic, pwns the enterprise Linux sector and employs the majority of kernel hackers (and just ballsed up royally with its recent partnership - *Symantec*, for gods' sake! - but they should weather it ok). * Mandriva - still kicking, playing interesting tunes on 3D desktop usability. * Various other commercial appliance distros e.g. firewalls, Tivo, etc. * One lone idealistic guy with who owns the damn trademark.
So let's not moan doom & gloom too early, eh?
Now, if someone rings tomorrow to tell me that Torvalds just sold Linux(tm), then you might have a point. But the *source* will still be out there & owned by the community that developed it. There is now a minimum level of code & application quality that proprietary software houses must meet; and while they don't, there will always be an underdog.
> It occurs to me that I'll probably get marked Flamebait for that. Still, although I don't > like Microsoft's business practices I do believe that if you want to bash them, bash them > based on facts not bullshit.
Hmm, on this occasion he may have a point, tho'. I haven't had an MS product in my house for *8* years, but it doesn't stop me knowing MS systems & products more than any other user chez mon employer...I may not use Windows by choice, but that doesn't stop me knowing about it.
BTW, I agree; Outlook is a bloated, horrible application...yet its value to business seems to be woefully underestimated by the development community. Evolution is nice enough but no real replacement (although maybe I should upgrade my home PC from FC4 before I sound off too much about system-freezing memory leaks).
> The Linux Kernel Guru needs to go back and start guru'ing his kernel and stop telling us > what is good or bad in the industry from the stand point of patents.
And who are you?
Let me step back and get some perspective, here...
Alan Cox made himself a legend on the basis of sheer productivity. He's repeatedly tackled areas of the kernel that many others thought thankless, difficult, and frankly too damn intimidating. E.g. debugging & eventually taking over the original networking stack; developing the original multiprocessing code; introducing some semblance of order to the ATA driver mess. So yes, I candidly agree with your label of "Linux Kernel Guru"; I'm with you so far.
Alan Cox has for years been a point of contact for large corporates and governments, and has influenced everything from global banking IT systems through to government policy and antitrust legislation. He's campaigned tirelessly against the DMCA, software patents, and other abusive/restrictive legislation & infrastructure. You seem to think that's of little value; I'd say he's eminently qualified to volunteer comment on Linux strategy. But on this occasion, as usual, he didn't volunteer his opinion; he was *asked* for it.
Alan Cox is widely known to be a "regular guy", unaffected by whatever recognition and awards he's given, and generous to boot. I happen to know - and hopefully he won't be too annoyed at me for saying this publically - that the award he was quietly most touched by was the 2003 FSF Award for the Advancement of Free Software.
When was the last time you contributed to the kernel? When did you last communicate anything beyond a cheap abusive posting on/.? When can we expect to see you nominated for a FSF award?
You have manifestly exercised your right to free speech & expression; now let me reciprocate :
Stop wasting bandwidth and go and do something more useful instead, you pointless piece of white trash.
Mod parent down more, please. I'm really interested in seeing where this "obvious-bullshit vs do-not-feed-the-troll" curve will minimise.
> I guess my point was that when dealing with the retrospective population types of studies > you mentioned, dealing with confounding factors (socioeconomic status, healthy worker > effect) is a pain in the ass, and really hurt the statistical reliability of your results. > That said, regulatory agencies tend to go overboard in dismissing these studies based on the > confounding factors.
*Nod* If you're overzealous there comes a point where confounding factors can bring you to a complete standstill.
In many fields where the data is there for the taking, e.g. medicine, the trend almost seems to be "discard by default". In a safety-critical industry where you have to pay for testing (guess where I work), the default is to *never* discard data unless you can *prove* it isn't representative of the population.
I'm strictly agnostic on the question of distribution type, but the advantage of a gaussian (normal) distribution is that it should be able to tolerate a number of confounding factors as the sample size becomes very large. The confusion should all "average out" in the end, as long as care is taken to ensure that there is no systemic bias. The weibull family of distributions contain a close approximation to the normal.
Bias can be avoided by increasing the size of the data sample, using batches of data from a variety of other sources (e.g. areas of high, low, and average socioeconomic status). Also, a useful tool is the probability plot, i.e. a plot of CDF vs variate: this can help to identify possible groups of outliers that can be discarded after investigation into the cause of skewness.
If I was giving advice on the radiation study issue, I'd recommend simply expanding the scope. There's a whole planet out there full of inhabited areas of unusually high ambient radioactivity (no need to go around asking people to voluntarily ingest substances that could be harmful to their health)! Even if budget & other factors restrict a study to a localised area (e.g. a depressed uranium mining town) it should still be possible to compare the results with a huge sample of depressed towns (centred around either just mining/quarrying, or possibly a variety of industries), and to draw viable conclusions.
(BTW, while Cornwall is for the most part strictly rural, the attractive thing about it - if the local ambient radiation is sufficiently high to be of interest - is that the strong tourist industry pumps enough money into the county to give it a fair socioeconomic cross-section).
> > The dispute on effectiveness is irrelevant. > As a twice-deployed Army vet, I'd say that's a matter of opinion.
Performance clearly doesn't seem irrelevant to the armour wearer who's being shot at.:) My point, however, was that it's probable that in the Army's view weight considerations overrode all other factors, and that they could have saved themselves a lot of bad publicity if they'd stated this up-front.
> 19 pounds might seem like a lot to someone who's used to wearing, on a regular basis, perhaps 5 pounds of clothes; but > to someone who goes to work carrying 10 pounds of water, 8 pounds of ammo, 4 pounds of grenades, 10 pounds of radio > batteries, 4 pounds of helmet, 8 pounds of rifle, and about 10 pounds of other miscellaneous gear, the difference is > negotiable. Deciding between an 70 pound load and a 90 pound load is far less significant.
As you say, 19lbs is a lot to someone who wears 5lbs clothing. But, as you also point out, the average infantryman is already overloaded - "we drag our asses around carrying too much gear". A 20% weight increase is *plenty* enough to slow you up even further in a firefight, and if your vehicle gets knocked out you're completely shafted. Armour has to be able to deal with a wide range of scenarios.
If armour is so heavy that you can barely move, all it does is turn you into one big, very slow-moving target. And at that stage it doesn't matter how impressive the performance of your armour is, something's going to get through; perhaps some bright spark will even manage a head shot.
> It's well within a range > where tactical considerations should decide which armor to use, rather than a blanket decree from some fucktard West > Point grad commanding a battalion of paper pusher at the Pentagon.
Absolutely. The canned statement I provided in my original post was simply how, if I were the USArmy spokesperson, I would have presented the fucktard's decision.
There's a trade-off between armour and mobility; the ideal balance point will vary for different missions, and be influenced further by available budget. Although I think the Army probably made the right general call, I've yet to meet a blanket decree that I like in person. There's plenty of scope for some roles to use heavier armour; a general policy modified by tactical considerations would be better.
> > They had a golden opportunity to appear both professional and caring, and somehow ended up looking irrevocably like a > > bunch of corrupt, disingenuous wankers. > "Somehow"? They ended up looking that way because that's exactly what they are.
> Good, controlled research on the topic of low-dose radiation with respect to humans is hard to come by, > however, considering the major stigma attached to radiation. Other than worker and background radiation > studies, you've pretty much got to wait until an accident happens.
Good set of posts, shame you finished this on with such a line of crap.:)
1/ Cornwall, UK: higher-than-average radiation in the granite geology there. Medical records should go back quite a way: compare them with the rest of the UK. There will be a number of other geographical locations, say anywhere in the world that there's a uranium mine. (Somewhere in the former USSR there's even a subterranean town where nuclear waste used to be processed...saw a Channel 4 documentary on it a few years ago, which ended with some bloke drinking a glass of water from the pond where they "cool" the waste...)
2/ We don't need any *more* accidents; thanks to long isotope half-lives we can draw on the mistakes of the past. In particular: parachute into the centre of Chernobyl, start walking 'til you get to active population. The advantage of this approach is that you should be able to create nice graphs of exposure intenisty vs beneficial/deleterious effects.
Nice, big statistical samples. The one difference with these approaches is that the water table tends to be contaminated, too; in your radioactive steel buildings people won't be actually *ingesting* higher radioactivity in the same way.
But otherwise the data's there; it's just a question of grabbing it. And I can't believe that no-one's done this already.
"It's simply a question of weight. Standard armour weighs 28lbs; Dragonskin weighs 47lbs. Despite increased flexibility and arguably better protection, our people can't operate effectively or safely under that penalty - their mobility, stamina, ordnance, and other equipment would all be unacceptably reduced. Therefore, with regret, the USArmy cannot allow the use of Dragonskin for military personnel. However we concede that it *may* be of use to noncombatants, for example news reporters."
USArmy public relations are a brain-dead bunch of morons for not being more up-front about it. They had a golden opportunity to appear both professional and caring, and somehow ended up looking irrevocably like a bunch of corrupt, disingenous wankers. It's a classic example of the type of autistic military "communications" that have made the USA so reviled in the Middle East & elsewhere.
Much respect to soldiers on the groud - the US jarhead, the UK marine with sore buttock, whoever. But USArmy high command should f*ck off, re-read their copies of Sun Tzu, and notice the bits that stress the importance of diplomacy in war.
I've often wondered along the same lines, and concluded that the "liquid water" criterion is the result of two overarching concerns:
1/ Communication. The most commonly quoted reason is that there's little point in looking for life-as-we-DON'T-know-it, as we'll be hard put to ever communicate with it.
Successful communication requires common frames of reference. Psychology is largely determined by environment; we're best equipped to communicate with creatures of roughly the same size as us (i.e. or a mouse, or elephant; no protazoa or magellanic gas clouds), who think roughly in the same terms as us, and therefore who must experience vaguely the same group of sensations (and emotions?) as us. In other words, life-as-we-know-it, which is by definition carbon-based lifeforms in an environment that contains large amounts of liquid water.
2/ Colonisation. While asteroid mining is likely to be a far more economical means of obtaining raw natural resources, in the long term a planet offers huge potential for expansion, and the possibility of easing the population pressure on Earth's environment...besides which wanderlust is just plain bred into us.
To survive, a human colony at that distance *must* be self-sufficient in terms of consumables. The focus on liquid water provides immediate confirmation of a couple of criteria:
i) Water is vital to terrestrial life. It's also a type of "sequestered oxygen" from which the breatheable stuff can be electrolysed and pumped into atmospheric domes if a human-sustaining atmosphere isn't already available.
ii) If liquid water exists it suggests that on at least some of the planetary surface, temperatures can probably support a terrestrial ecosystem with a minimal need for containment. Building a series of sealed domes to live in is fairly easy & sustainable, and suggests terraforming is possible; heating and/or cooling a sealed dome is far too expensive & error-prone to survive for long.
Hence if liquid water exists on a planet, there's an outside chance we can settle there, or at least talk with whoever does. Without liquid water the possibilities are much smaller, and likely to be less sustainable and economic even if the damn place is made of solid gold with pure crude oil oceans.
It's a game of chance. The search for liquid water just maximises the probability of a pay-out.
"Great post but kinda hard to read."
Yes, sorry about that. My first Slashdot post; if I'd hit "Preview" I'd have noticed that all my formatting had disappeared.
Cheers,
C
Strictly it doesn't make sense, that's why.
The premise is sound - they plot the sizes of all dinosaurs, and find the trend follows a single bell curve instead of a series of spikes. Since they're dealing with clones there aren't a lot of reasons for big variations in dino size. Because there's a single, broad population instead of a number of narrow ones then they have either one single batch (which they know not to be true) or many, many more batches than they thought...so many that the individual spikes blend into a whole, in fact. Possible conclusions - either many more batches are being created than are recorded (which can be easily checked via their own internal investigation), or (essentially the same thing) the dinos are breeding.
Now for the fast-and-loose bits, i.e. why this doesn't make sense.
1/ Most of the dinos couldn't breed when released, but a few or possibly all managed to start doing so owing to a vaguely-mentioned bit of trans-species DNA splicing (from modern frogs, if I recall, but since I haven't read the novel since it was first released in the UK, ~2 years b4 film, I can't be sure). Therefore there would still be signs of more than one population; first a couple of spikes, then the start of a kind of "spiky hump" which would subsequently merge into a smooth bell at the extreme tail. *That* kind of behaviour is unmistakeable, and would raise instant alarms.
2/ Not extremely relevant here, but it's something of a personal crusade for me: it's a common fault to spot a bell curve and assume it's "Normal". Note that a normal gaussian distribution is asymptotic in both directions. While I'm prepared to accept for the sake of argument that a dino 5 times the size of its siblings wouldn't realistically happen because the probability is too remote, I can't accept that a dino could ever be born with *negative* height; and assuming a single normal distribution allows for this possibility. Either the normal distribution should be applied to the *logarithm* of the dino dimensions (i.e. the so-called log-normal distribution, which will pin the minimum dino size to zero) or another distribution form should be tried. The Weibull family of distributions shows good correlation with life & size data, and includes a fair approximation to the normal & lognormal (albeit with fatter tails and a minimum life/size parameter).
Overall, though, I loved the book because of its overall authenticity; and I was kind of sad that I could only resort to statistics to judge the veracity of the chaos maths arguments that underpin the storytelling (my understanding of choas maths stops at "Ooh, pretty fractals!").
The film, however, ranked at the top of my most hated of all time until the release of "I, Robot". There's a pattern: I was excited about the release of both films because of the hard science, intricate storytelling, and characters & character-deaths (the mathematician in JP is just so hilariously nihilistic). The films completely decimated both (what the hell happened to Susan Calvin & the refutation of the Frankenstein complex?), in the interests of effects and general "Hollywoodising" to pull in the crowds.
The question was to list good, hard science sci-fi literature. Here's a TV/movie media example: Firefly/Serenity. Most things in those was spot-on - no sound in space; projectile firearms still useful; laser weapons with short battery lives and with quiet, sizzling beams that cut, not loud "bolts" that punch; military cruisers without a hint of streamlining; no aliens or AI to speak of; dying of cold before asphyxiation if your life support cuts out; random (and heartbreaking!) character deaths.
That kind of detail is the kind of thing usually ditched & glossed over in the interests of general appeal, yet those details were half the reason I was captivated from the first episode. Kudos to the scriptwriting, acting, FX, etc., but the cultures, languages, and technical accuracy is what made me believe I was actually seeing it all happen. Brilliant.
Slashdot Mirror
Ok, I give up. How does one moderate the *original*article* as flamebait?
C
To all those bitching about how a $50 discount won't convert any Windows users: you're missing the point.
Dell is in shit, in a very competitive, saturated market. Common business sense says the best thing they can do is to be the first to into a new market.
Headline: the new market is not one of potential converts from Windows - not yet, anyhow.
Dell's hoped-for market consists of :
* long-term linux users with Real Jobs who need a new PC but don't have the time or just can't be arsed to build, install, and customise Yet Another Linux Box.
* The type who sometimes, against their best wishes, might feel themselves lured back to Windows because they feel it's just *easier*.
* The type who would quite like official support on a preinstalled linux box, instead of banging their heads against the ITSupport wall when the Windows box they just bought, wiped, reinstalled & voided the warranty for has a hardware problem.
These types are likely to discriminate in the long term against a company that quoted them *more* for a preinstalled linux box. For these types, $50 is not an outright incentive, but it "is" a nice goodwill gesture. Converting Windows users comes later.
Ladies and gentlemen, Dell's intended market is *us*. And no-one's bothered with us before.
As the man said elsewhere : Vote. Wallet. Now. I know I will as soon as the UK gets the same offer.
Cheers,
C
> Since this seems like such a new concept (please correct me if I'm wrong; I
:
> don't follow plane technology too much), it would just seem prudent to try
> bending the wings until they break...
There's lots of reasons not to test a CFRP wing to failure.
1/ Wing test rigs are designed to break standard metallic wings. Rigs of this scale are *not* designed for wings that deform this much - some fancy rig redesign would be necessary to keep the loading plane normal on a wing that can double back on itself!
2/ Aerostructural failure is already defined by the limit load & ultimate load criteria. Basically, limit load is defined by certification authority book cases, and is the maximum load the aircraft would be expected to see in service, plus some. There can be no detrimental deformation at limit load. Above limit load, separation of structure must not occur, although repair may be required. Ultimate load is the criterion defined as 150% of limit load (for civil aircraft - it's different for military); if an a/c is ever subjected to ultimate load then it should be retired. Therefore there's no "prudence" in testing a wing (much) above ultimate load; the only purpose it would serve is to check if your calcs are overly conservative and/or your manufacturing & assembly is up to scratch.
3/ A standard a/c design (such as every civil airliner currently on Earth) would fall out of the sky for aerodynamic reasons long, long before its wingtips touched. In fact, it's possible that much of the wing would have to be overstrength just to maintain the stiffness required for aerodynamic purposes.
4/ As pointed out elsewhere in this thread, in metallic structure energy is dissipated as plastic deformation (& heat); carbon fibre is almost purely elastic, hence *none* of the energy is lost until fracture. I wouldn't want to be in the same room as that large a chunk of CFRP containing enough energy to bend a wing through 180 degrees (frankly I wouldn't even want to be on the clean-up crew afterwards - all that dust)! This is the same reason that metallic pressure vessels in chemical works are tested hydraulically, not pneumatically (gas = compressive = 1 big spring).
5/ The criterion is that any single component has how can they make accurate judgments and calculations without knowing exactly > how much stress the wings can take before snapping?
Point 3/ explains why a specific wing's breaking limit is irrelevant.
However, you kind of have a point in terms of correlation of calculation methods with known failure, especially because composite technologies are so new, unknown, and CFRP structure may be (probably is) overdesigned.
The reality of overtesting a composite wing is much more complex, though.
Yes, continuing a static test 'til failure would provide some useful feedback on analysis techniques. Not as much as you'd think, though
a/ Even in metallic structure there's a lot of scatter associated with material properties, assembly, etc. There are several detail areas in aircraft design where the calcs are known to be conservative: although removing conservatism can increase a/c efficiency, it is ultimately more important to design robustly if a detail is very sensitive to small changes in local load.
b/ Composite strength is affected by a huge number of issues that don't affect metals. Manufacturing & assembly errors can cause problems (in effect you create the material at the same time as you make your structure). Moisture ingress into the matrix can affect material strength considerably and also add weight; prolonged exposure to fuel will also degrade the material. 2D CFRP laminates are very intolerant to contact and impact, so birdstrike or damage from runway debris will also weaken the material. To summarise, there are a number of new factors, all have to be accounted for, which leads to a wing that is several times stronger than it needs to be, simply to ensure that it's *still* strong enough on its
Good point.
One error. In the interests of accuracy :
> Additionally, it was determined that the concrete foundations were binding
> CO2 as they cured (concrete cures for years and years), causing still more
> problems"
No. Dangerously high levels of CO2 terminated the biosphere2 project prematurely.
The cause was a naive assumption during dome construction that there was no upper limit to the degree of soil fertility in a sealed environment: when preparing the soil, it was decided to boost the vegetation (hence O2 levels) by digging in lots of extra organic matter - I think they tripled the originally-quoted quantity. The slight increase in O2 output was more than offset by skyrocketing CO2 levels from the decaying organics. The inhabitants suffered severe CO2 poisoning, could barely continue functioning, and eventually had to crack the seal.
Subsequent investigations showed a large mass of CO2 was unaccounted for in the dome's atmosphere. That's when someone realised that the strange chemical changes that had been observed in the concrete structure had actually been sequestering the CO2; the concrete was slowly turning to something like limestone.
Too much organic matter doomed the project to failure. The concrete structure actually *extended* its life considerably. As far as I know, no studies were conducted to examine the long-term effects on the viability of the structure.
C
I can't agree with the implied criticism of "MirrorMask". I simply loved it, so did my wife.
Sure it lacked mass appeal; it was very much an "art-house" feel, and the cast of only 7 or so people ensured that the film felt more like a screenplay (not necessarily a bad thing in my opinion - check out "Dogville").
But the plot was inventive and quintessentially Gaiman, and the acting was top notch. Costumes and sets were simple but well thought out, and seeing Dave McKean's usual style of artwork transformed from the static page to cgi was absolutely jaw-dropping; it felt like 3D, somehow.
Weirder, both the wife and I had really powerful, lucid dreams that night. How many blockbusters can you say *that* about?
Original.
Conrad
Typical other responses to the /. posting :
"Mr. Shuttleworth is to be applauded for not bending to Microsoft."
"It's pleasing to hear that he will not be directly working with Microsoft"
"Dear Ballmer. NUTS. Signed, Mark Shuttleworth"
Ok, so Shuttleworth hasn't taken MS's shilling on the patent issue: but he hasn't exactly given a "Nuts" reply either, and has most certainly not discounted directly working with MS, quite the opposite.
From Shuttleworth's blog:
1/ "We have declined to discuss any agreement with Microsoft under the threat of unspecified patent infringements,"
2/ "Allegations of 'infringement of unspecified patents' carry no weight whatsoever. We don't think they have any legal merit, and they are no incentive for us to work with Microsoft on any of the wonderful things we could do together. A promise by Microsoft not to sue for infringement of unspecified patents has no value at all and is not worth paying for."
3/ "I have no objections to working with Microsoft in ways that further the cause of free software," "and I don't rule out any collaboration with them, in the event that they adopt a position of constructive engagement with the free software community."
I read this as :
1/ "We *have* been approached Microsoft, who insisted on the patent clause as part of the deal" .
2/ "Just in case MS thinks that our refusal is just us stalling for a higher price, let me make it clear: pull the other one, guys, it's got bells on it. *Any* price you offer will not reimburse the value lost to us if we publically state that MS has IP in Linux OS".
3/ "If MS really *are* for real on the question of interoperability, then we're still listening.
Shuttleworth is refusing to capitulate on the patent issue, because if he does he signs away the value of his company. He calls MS' bluff by implying that Canonical would insist on opening the relevant interoperability source, but says it in a way that if MS are for real, then no-one's lost face and negotiations can proceed.
My estimation of Mr Shuttleworth's business acumen just shot up.
Cheers,
C
> > Personally I detest the proprietary shit its peddling, but Ubuntu's already proven there's
> > a demand for that.
> Ubuntu went so far as to create a "restricted driver manager" that tells you when you're
> using binary drivers, why you shouldn't, and what you can do about it.
You appear to be confused.
I didn't criticise Ubuntu. I actually quite *like* Ubuntu's pragmatic take on proprietary drivers:
* They're in the distro because they work, and there's currently no other option other than breaking functionality.
* This helps win Windows users over, who want their systems to Just Work (tm). Linux user base grows.
* New users are then educated by the restricted driver manager.
* Completely free version of Ubuntu is in the pipeline.
I think this is a smart way forward, and Ubuntu will be my 2nd choice of distro if Fedora ever goes seriously downhill or bites the dust.
I *still* hate the proprietary shit. Have done for years, ever since I updated my kernel, yum couldn't find a matching NVidia 3D binary driver, and my next reboot left me at a text console. Can't wait 'til Nouveau is ready for the mainstream.
> If you call this "peddling proprietary shit", then I don't think you understand more than
> one of those words.
I understand just fine: you misread my post, and thought I was criticising what's obviously favourite distro (Ubuntu) when I was actually criticising CNR. So you then posted a heated personal reply.
Like Ubuntu: *hate* proprietary drivers & codecs. Is that clear enough?
Conrad
> Many years ago Microsoft started to hire all the famous people in free software. Soul after
> soul joined the dark side and the future looked grim.
Hardly. The founder of Gentoo went to MS...and got out again quick, if I remember correctly. What other high-profile defections have there been?
> Who do you reakon will be the next distro to enter into a Microsoft deal. So far its been
> all the ones you'd expect - I suspect we won't be hearing much more of Xandros and Linspire,
> although I think Novell could still come out unscathed- My guess is Mandriva - although it
> saddens me to say it - they are having financial troubles etc.
I suspect you're right. But then again the strong French socialist streak may win through.
> Mark Shuttleworth said he wouldn't go into any kind of deal like this and I think I believe
> him.
I definitely do - it would be a travesty of all the reasons he says he got into the business, and he's not exactly strapped for beer money.
> Most of all we've got to hope that Red Hat doesn't,
If they do then it will be a pyrrhic victory for MS; I suspect the other distros would shortly be welcoming a lot of new kernel developers into their ranks. Or we might see the start of another rpm-based Linux distro. What's the value of a Linux company without its engineers/developers?
Besides, MS don't *need* to take out Redhat. By turning Novell they already threw a spanner in the works for the Linux Standards Base; if Linux is already fragmenting why draw any additional antitrust heat?
> as the largest commercial Linux company it would be disastrous for any possible defence we
> have against possible patent issues etc
Not really.
Redhat has already released virtually all their code - CentOS is a good showcase for what's left when you strip out everything proprietary.
As to patent issues, who at Redhat do you think chases those? The business guys? Really?
Clue: ever wondered why Alan Cox (FSF Award winner 2003, anti-s/w patent activist) isn't as code-productive as he was a few years ago?
In a market where there is an infinite supply of the raw product (e.g. baked beans), you have to compete on "value added", which often manifests a perceived quality brand. In the software world the engineers & developers are what add value.
Any linux distro that loses its developers is dead in the water - check out SCO.
Put it another way. You're worrying too much about the corporates; keep your eye on their staff instead.
Cheers,
Conrad
> ...continues to hang itself. At least we still have Debian. Even though its derivatives will > probably all sell out.
A little perspective here, please.
* Novell sold out because, despite their purchases of Ximian & SuSE, they never really "got" Linux; they were just trying to shore up a rapidly dying Netware product while continuing business along the same paradigms that killed it. Witness the resulting exodus of several core SUSE developers, further reducing the company's understanding of Linux; frankly I've seen Novell Linux brands as almost defunct for some years now. (go on, flame away...)
* Xandros sold out because their market share and community was miniscule. They sought to steal Windows market share, but (unsurprisingly) didn't have the resource to tackle Redmond. Xandros are already defunct and starting to smell; they just don't know it yet. (go on, flame some more...)
* Linspire haven't really recovered since having their teeth pulled, and they really don't "get" the security issue. The whole distro is very much Kevin Carmony's baby, and seems to be very fluid while it tries to find a profitable niche. Ubuntu's just broken into the territory it was trying to win (i.e. preinstalled mainstream linux), so I think the distro will die soon. Strangely, though, I don't think that Linspire has sold out, exactly, it's following in its father's footsteps; it understands business, not OS, and is evolving into a kind of "software accessibility enabler". Personally I detest the proprietary shit its peddling, but Ubuntu's already proven there's a demand for that.
So MS has munched on the low-hanging fruit. Sad, but not unexpected; the old & weak are always the first to go in war & business. What remains is :
* Several hundred non-commercial distros, top of the list is Debian, the epitome of idealism.
* Ubuntu - very smart, idealistic, breaking into the mainstream.
* Redhat - very smart, idealistic, pwns the enterprise Linux sector and employs the majority of kernel hackers (and just ballsed up royally with its recent partnership - *Symantec*, for gods' sake! - but they should weather it ok).
* Mandriva - still kicking, playing interesting tunes on 3D desktop usability.
* Various other commercial appliance distros e.g. firewalls, Tivo, etc.
* One lone idealistic guy with who owns the damn trademark.
So let's not moan doom & gloom too early, eh?
Now, if someone rings tomorrow to tell me that Torvalds just sold Linux(tm), then you might have a point. But the *source* will still be out there & owned by the community that developed it. There is now a minimum level of code & application quality that proprietary software houses must meet; and while they don't, there will always be an underdog.
Best regards,
Conrad
> It occurs to me that I'll probably get marked Flamebait for that. Still, although I don't
> like Microsoft's business practices I do believe that if you want to bash them, bash them
> based on facts not bullshit.
Hmm, on this occasion he may have a point, tho'. I haven't had an MS product in my house for *8* years, but it doesn't stop me knowing MS systems & products more than any other user chez mon employer...I may not use Windows by choice, but that doesn't stop me knowing about it.
BTW, I agree; Outlook is a bloated, horrible application...yet its value to business seems to be woefully underestimated by the development community. Evolution is nice enough but no real replacement (although maybe I should upgrade my home PC from FC4 before I sound off too much about system-freezing memory leaks).
Cheers,
C
> Alan Cox is full of shit.
/.? When can we expect to see you nominated for a FSF award?
> The Linux Kernel Guru needs to go back and start guru'ing his kernel and stop telling us
> what is good or bad in the industry from the stand point of patents.
And who are you?
Let me step back and get some perspective, here...
Alan Cox made himself a legend on the basis of sheer productivity. He's repeatedly tackled areas of the kernel that many others thought thankless, difficult, and frankly too damn intimidating. E.g. debugging & eventually taking over the original networking stack; developing the original multiprocessing code; introducing some semblance of order to the ATA driver mess. So yes, I candidly agree with your label of "Linux Kernel Guru"; I'm with you so far.
Alan Cox has for years been a point of contact for large corporates and governments, and has influenced everything from global banking IT systems through to government policy and antitrust legislation. He's campaigned tirelessly against the DMCA, software patents, and other abusive/restrictive legislation & infrastructure. You seem to think that's of little value; I'd say he's eminently qualified to volunteer comment on Linux strategy. But on this occasion, as usual, he didn't volunteer his opinion; he was *asked* for it.
Alan Cox is widely known to be a "regular guy", unaffected by whatever recognition and awards he's given, and generous to boot. I happen to know - and hopefully he won't be too annoyed at me for saying this publically - that the award he was quietly most touched by was the 2003 FSF Award for the Advancement of Free Software.
When was the last time you contributed to the kernel? When did you last communicate anything beyond a cheap abusive posting on
You have manifestly exercised your right to free speech & expression; now let me reciprocate :
Stop wasting bandwidth and go and do something more useful instead, you pointless piece of white trash.
Mod parent down more, please. I'm really interested in seeing where this "obvious-bullshit vs do-not-feed-the-troll" curve will minimise.
Cheers,
Conrad
*wry smile*
Ok, that kind of scratches my opinions.
Nice research, btw. Can you point me to TFA?
> I guess my point was that when dealing with the retrospective population types of studies
> you mentioned, dealing with confounding factors (socioeconomic status, healthy worker
> effect) is a pain in the ass, and really hurt the statistical reliability of your results.
> That said, regulatory agencies tend to go overboard in dismissing these studies based on the
> confounding factors.
*Nod* If you're overzealous there comes a point where confounding factors can bring you to a complete standstill.
In many fields where the data is there for the taking, e.g. medicine, the trend almost seems to be "discard by default". In a safety-critical industry where you have to pay for testing (guess where I work), the default is to *never* discard data unless you can *prove* it isn't representative of the population.
I'm strictly agnostic on the question of distribution type, but the advantage of a gaussian (normal) distribution is that it should be able to tolerate a number of confounding factors as the sample size becomes very large. The confusion should all "average out" in the end, as long as care is taken to ensure that there is no systemic bias. The weibull family of distributions contain a close approximation to the normal.
Bias can be avoided by increasing the size of the data sample, using batches of data from a variety of other sources (e.g. areas of high, low, and average socioeconomic status). Also, a useful tool is the probability plot, i.e. a plot of CDF vs variate: this can help to identify possible groups of outliers that can be discarded after investigation into the cause of skewness.
If I was giving advice on the radiation study issue, I'd recommend simply expanding the scope. There's a whole planet out there full of inhabited areas of unusually high ambient radioactivity (no need to go around asking people to voluntarily ingest substances that could be harmful to their health)! Even if budget & other factors restrict a study to a localised area (e.g. a depressed uranium mining town) it should still be possible to compare the results with a huge sample of depressed towns (centred around either just mining/quarrying, or possibly a variety of industries), and to draw viable conclusions.
(BTW, while Cornwall is for the most part strictly rural, the attractive thing about it - if the local ambient radiation is sufficiently high to be of interest - is that the strong tourist industry pumps enough money into the county to give it a fair socioeconomic cross-section).
> > The dispute on effectiveness is irrelevant.
:)
> As a twice-deployed Army vet, I'd say that's a matter of opinion.
Performance clearly doesn't seem irrelevant to the armour wearer who's being shot at.
My point, however, was that it's probable that in the Army's view weight considerations overrode all other factors, and that they could have saved themselves a lot of bad publicity if they'd stated this up-front.
> 19 pounds might seem like a lot to someone who's used to wearing, on a regular basis, perhaps 5 pounds of clothes; but
> to someone who goes to work carrying 10 pounds of water, 8 pounds of ammo, 4 pounds of grenades, 10 pounds of radio
> batteries, 4 pounds of helmet, 8 pounds of rifle, and about 10 pounds of other miscellaneous gear, the difference is
> negotiable. Deciding between an 70 pound load and a 90 pound load is far less significant.
10+8+4+10+4+8+10+47 = 101 lbs
10+8+4+10+4+8+10+28 = 82 lbs
As you say, 19lbs is a lot to someone who wears 5lbs clothing. But, as you also point out, the average infantryman is already overloaded - "we drag our asses around carrying too much gear". A 20% weight increase is *plenty* enough to slow you up even further in a firefight, and if your vehicle gets knocked out you're completely shafted. Armour has to be able to deal with a wide range of scenarios.
If armour is so heavy that you can barely move, all it does is turn you into one big, very slow-moving target. And at that stage it doesn't matter how impressive the performance of your armour is, something's going to get through; perhaps some bright spark will even manage a head shot.
> It's well within a range
> where tactical considerations should decide which armor to use, rather than a blanket decree from some fucktard West
> Point grad commanding a battalion of paper pusher at the Pentagon.
Absolutely. The canned statement I provided in my original post was simply how, if I were the USArmy spokesperson, I would have presented the fucktard's decision.
There's a trade-off between armour and mobility; the ideal balance point will vary for different missions, and be influenced further by available budget. Although I think the Army probably made the right general call, I've yet to meet a blanket decree that I like in person. There's plenty of scope for some roles to use heavier armour; a general policy modified by tactical considerations would be better.
> > They had a golden opportunity to appear both professional and caring, and somehow ended up looking irrevocably like a
> > bunch of corrupt, disingenuous wankers.
> "Somehow"? They ended up looking that way because that's exactly what they are.
*smile*
C
> Good, controlled research on the topic of low-dose radiation with respect to humans is hard to come by,
:)
> however, considering the major stigma attached to radiation. Other than worker and background radiation
> studies, you've pretty much got to wait until an accident happens.
Good set of posts, shame you finished this on with such a line of crap.
1/ Cornwall, UK: higher-than-average radiation in the granite geology there. Medical records should go back quite a way: compare them with the rest of the UK. There will be a number of other geographical locations, say anywhere in the world that there's a uranium mine. (Somewhere in the former USSR there's even a subterranean town where nuclear waste used to be processed...saw a Channel 4 documentary on it a few years ago, which ended with some bloke drinking a glass of water from the pond where they "cool" the waste...)
2/ We don't need any *more* accidents; thanks to long isotope half-lives we can draw on the mistakes of the past. In particular: parachute into the centre of Chernobyl, start walking 'til you get to active population. The advantage of this approach is that you should be able to create nice graphs of exposure intenisty vs beneficial/deleterious effects.
Nice, big statistical samples. The one difference with these approaches is that the water table tends to be contaminated, too; in your radioactive steel buildings people won't be actually *ingesting* higher radioactivity in the same way.
But otherwise the data's there; it's just a question of grabbing it. And I can't believe that no-one's done this already.
Cheers,
Conrad
The dispute on effectiveness is irrelevant.
I R'dTFA. My suggestion :
"It's simply a question of weight. Standard armour weighs 28lbs; Dragonskin weighs 47lbs. Despite increased flexibility and arguably better protection, our people can't operate effectively or safely under that penalty - their mobility, stamina, ordnance, and other equipment would all be unacceptably reduced. Therefore, with regret, the USArmy cannot allow the use of Dragonskin for military personnel. However we concede that it *may* be of use to noncombatants, for example news reporters."
USArmy public relations are a brain-dead bunch of morons for not being more up-front about it. They had a golden opportunity to appear both professional and caring, and somehow ended up looking irrevocably like a bunch of corrupt, disingenous wankers. It's a classic example of the type of autistic military "communications" that have made the USA so reviled in the Middle East & elsewhere.
Much respect to soldiers on the groud - the US jarhead, the UK marine with sore buttock, whoever. But USArmy high command should f*ck off, re-read their copies of Sun Tzu, and notice the bits that stress the importance of diplomacy in war.
Conrad
I've often wondered along the same lines, and concluded that the "liquid water" criterion is the result of two overarching concerns:
1/ Communication. The most commonly quoted reason is that there's little point in looking for life-as-we-DON'T-know-it, as we'll be hard put to ever communicate with it.
Successful communication requires common frames of reference. Psychology is largely determined by environment; we're best equipped to communicate with creatures of roughly the same size as us (i.e. or a mouse, or elephant; no protazoa or magellanic gas clouds), who think roughly in the same terms as us, and therefore who must experience vaguely the same group of sensations (and emotions?) as us. In other words, life-as-we-know-it, which is by definition carbon-based lifeforms in an environment that contains large amounts of liquid water.
2/ Colonisation. While asteroid mining is likely to be a far more economical means of obtaining raw natural resources, in the long term a planet offers huge potential for expansion, and the possibility of easing the population pressure on Earth's environment...besides which wanderlust is just plain bred into us.
To survive, a human colony at that distance *must* be self-sufficient in terms of consumables. The focus on liquid water provides immediate confirmation of a couple of criteria:
i) Water is vital to terrestrial life. It's also a type of "sequestered oxygen" from which the breatheable stuff can be electrolysed and pumped into atmospheric domes if a human-sustaining atmosphere isn't already available.
ii) If liquid water exists it suggests that on at least some of the planetary surface, temperatures can probably support a terrestrial ecosystem with a minimal need for containment. Building a series of sealed domes to live in is fairly easy & sustainable, and suggests terraforming is possible; heating and/or cooling a sealed dome is far too expensive & error-prone to survive for long.
Hence if liquid water exists on a planet, there's an outside chance we can settle there, or at least talk with whoever does. Without liquid water the possibilities are much smaller, and likely to be less sustainable and economic even if the damn place is made of solid gold with pure crude oil oceans.
It's a game of chance. The search for liquid water just maximises the probability of a pay-out.
Cheers,
C
And again!
Formatting, i.e. paragraph separation, disappears after I preview. What's going on?
Will try POT instead of the default HTML setting.
C
"Great post but kinda hard to read." Yes, sorry about that. My first Slashdot post; if I'd hit "Preview" I'd have noticed that all my formatting had disappeared. Cheers, C
Strictly it doesn't make sense, that's why. The premise is sound - they plot the sizes of all dinosaurs, and find the trend follows a single bell curve instead of a series of spikes. Since they're dealing with clones there aren't a lot of reasons for big variations in dino size. Because there's a single, broad population instead of a number of narrow ones then they have either one single batch (which they know not to be true) or many, many more batches than they thought...so many that the individual spikes blend into a whole, in fact. Possible conclusions - either many more batches are being created than are recorded (which can be easily checked via their own internal investigation), or (essentially the same thing) the dinos are breeding. Now for the fast-and-loose bits, i.e. why this doesn't make sense. 1/ Most of the dinos couldn't breed when released, but a few or possibly all managed to start doing so owing to a vaguely-mentioned bit of trans-species DNA splicing (from modern frogs, if I recall, but since I haven't read the novel since it was first released in the UK, ~2 years b4 film, I can't be sure). Therefore there would still be signs of more than one population; first a couple of spikes, then the start of a kind of "spiky hump" which would subsequently merge into a smooth bell at the extreme tail. *That* kind of behaviour is unmistakeable, and would raise instant alarms. 2/ Not extremely relevant here, but it's something of a personal crusade for me: it's a common fault to spot a bell curve and assume it's "Normal". Note that a normal gaussian distribution is asymptotic in both directions. While I'm prepared to accept for the sake of argument that a dino 5 times the size of its siblings wouldn't realistically happen because the probability is too remote, I can't accept that a dino could ever be born with *negative* height; and assuming a single normal distribution allows for this possibility. Either the normal distribution should be applied to the *logarithm* of the dino dimensions (i.e. the so-called log-normal distribution, which will pin the minimum dino size to zero) or another distribution form should be tried. The Weibull family of distributions shows good correlation with life & size data, and includes a fair approximation to the normal & lognormal (albeit with fatter tails and a minimum life/size parameter). Overall, though, I loved the book because of its overall authenticity; and I was kind of sad that I could only resort to statistics to judge the veracity of the chaos maths arguments that underpin the storytelling (my understanding of choas maths stops at "Ooh, pretty fractals!"). The film, however, ranked at the top of my most hated of all time until the release of "I, Robot". There's a pattern: I was excited about the release of both films because of the hard science, intricate storytelling, and characters & character-deaths (the mathematician in JP is just so hilariously nihilistic). The films completely decimated both (what the hell happened to Susan Calvin & the refutation of the Frankenstein complex?), in the interests of effects and general "Hollywoodising" to pull in the crowds. The question was to list good, hard science sci-fi literature. Here's a TV/movie media example: Firefly/Serenity. Most things in those was spot-on - no sound in space; projectile firearms still useful; laser weapons with short battery lives and with quiet, sizzling beams that cut, not loud "bolts" that punch; military cruisers without a hint of streamlining; no aliens or AI to speak of; dying of cold before asphyxiation if your life support cuts out; random (and heartbreaking!) character deaths. That kind of detail is the kind of thing usually ditched & glossed over in the interests of general appeal, yet those details were half the reason I was captivated from the first episode. Kudos to the scriptwriting, acting, FX, etc., but the cultures, languages, and technical accuracy is what made me believe I was actually seeing it all happen. Brilliant.