I was always under the impression that patents were for inventions and not ideas
That's right.
[...] (minus this crap about software patents).
Software is only patentable when it is an "invention". For a long time it wasn't patentable at all. If I recall correctly, the logjam broke when somebody snuck one through by describing a hardware device to do the same computation, then griped about it in court. And another patented an invention that included a software-driven component.
IMHO a piece of software is simultaneously a "mathematical algorithm" and a "work of composition" but not a "performance" (even if it DOES prescribe a machine's actions). So it should be unpatentable, the source code should be copyrightable (though the current near-infinite copyrights are nuts), and the copyright should carry over to the object as a "derived work". And it should be subject to reverse-engineering of how it works, fair use of tiny snippets of code, but not verbatim plagarism of the whole or large chunks of code.
What if you write some vague description of a computer program or some device, but can't figure out how to make it work with all the details in it? That's where the problem comes in.
Indeed. But part of the requirements that the patent examiner is checking is that the description is adequate for "someone expert in the art" to "perform the invention". Sometimes it's complicated enough, or enough beyond the examiner's expertese, that the examiner (who has limited time) lets it go through even if it really is broken or incompelete. (And that's how the occasional perpetual motion machine still gets patented.) But
They should require working models of the patented products (in the case of software: working source code if there really must be software patents)) so that you're sure that the person patenting the invention has done the real work of inventing something.
Why? They DID that. And they got BURIED in working models.
What do you do when making the first one costs billiions? Say: An orbital skyhook design? Or how about the "whip" satelite launcher? (Trail the satelite behind a 747 on a long cord, do a loop and "crack the whip", releasing the satelite when it's headed beyond the atmosphere and the 747 is flying a couple miles lower.) Are you saying deliver a skyhook or a 747 with a whip to the patent office or forego the patent?
What about a gene-engineered bacterium that kills colon cancer? A "working model" would include the patients of the human trials.
And the time to patent a lot of stuff is when it's still in the design stages - when you know WHAT you're doing but haven't had a hundred-man team dot the Is and cross the Ts of the implementation. (And three of 'em quit and go work for the competition.) "Working source code" is right out in such a situation.
One takes a look at all the patents given in the US that don't and can NEVER be produced because they are physically impossible.
It's not the job of the patent office to determine whether the invention actually works, or can even be constructed.
It's the job of the patent office to certify that, within the results of a reasonable search, what you registered is not already registered.
This certification is a time-limited "license to sue" anyone else who does the same thing the same way - and a presumption that you have precedence. (i.e. it's up to the other guy to prove that his thing is different, or that he had it before you did.)
A patent on something impossible is just a number used up in an aleph-null namespace, a few documents on file in a bureaucracy, and some manhours that were paid for by the guy who patented the impossible thing. In theory, nobody is going to infringe a patent-on-the-impossible, so no court case (unless the fruitcake starts suing people randomly, of course).
Single exception: Perpetual motion machines. The patent office was SO clogged with perpetual motion applications in the steam-engine era that they bogged down. So they still require a working model. (As, back when they started up, they did for everthing, before the office became clogged with working models. Which is where the Smithsonian got a lot of their exhibits. B-) )
Routinely bouncing putative perpetual-motion devices sometimes is a problem. For instance, an inventor came up with a VERY efficient still: Very tall, so the top is at a near-vacuum due to the weight of the liquids. Counter-current heat excanger transfers heat from the the condensate and brine to the feedstock, so most of the heat is recycled and the product and waste come out cool. Base-of-couumn pressure difference between the feedstock and the outputs, to create the temperature difference across the heat exchanger.
Patent office bounced it for being a perpetual-motion mcahine. Inventor got his patent after proving that you still had to energy to provide the heat-of-solution of the impurities and to replace losses in the (necessarily) less-than-perfect heat exchanger and through the insulation between the column and the environment. It was VERY GOOD - because the input heat got used many times rather than once - but it wasn't free.
Let's not misinterpret SCO's actions.
on
Linus on DRM
·
· Score: 1
It's nice to see that Linus has his priorities in the right place. Too bad others can't follow his example. *cough* SCO *cough*
Let's not misinterpret SCO's actions.
They're claiming that IBM took some of their licensed software and, in violation of the license, purported to sublicense rights they didn't have under the GPL.
Unlike the original UNIX (System N), whose proprietary status may be problematic, SVRn was carefully re-implemented under well-defined IP law. (In fact, some believe that was the whole POINT of SVRn.) It is NOT public unless its owner MAKES it public. Right now its owner is SCO.
If there really is proprietary code circulating with GPL copyleft attached, it needs to be identified, purged, and replaced - or otherwise made free.
For instance: IBM might counter-sue and the court might find that SCO had incorporated GPLed code with functionality of comparable value, and as part of the decision make the leaked SCO code public. (Courts DO like to upset the status-quo as little as possible if it can be done equitably.)
we all know you should wipe down after cleaning with whatever material you choose to apply. A spider web in the barrel of a shotgun can be lethal. There is always a need for care and common sense when dealing with any weapon.
Hear, hear.
I also note that the WD-40 faq you quoted does recommend it for guns - for protection from corrosion - and claims it won't harm bluing. But it doesn't recommend it for lubrication. The same site DOES recommend another of their products - 3-in-1 white lithium grease - for lubricating sliding parts on weather/dirt-exposed machinery - specifically including guns. (They also recommend their 3-in-1 silicone spray for protecting gun finishd surfaces from corrosion.)
And with Microsoft's latest effort to try to make their OS's as "secure" as possible, shouldn't all these people picking on opensource be targeting Microsoft as well, since they are now SECURE?
You mean the one where they give copies of the source code to the information warfare departments of all the major powers (including the US's former enemies - Russia and China), but still keep it a "secret" from the public-sector crypto and software scholars?
WD-40 is _not_ a thick oil. It has a carrier that evaporates, and is one of the thinnest, which is why it is sold as an aerosol. You're clearly confusing it with some other product.
My understanding is that WD-40 is a MIX of very light and very heavy fractions. The light fractions (which are the bulk of it) aid penetration and freeing of stuck parts (by disolving older tar-like stuff in the tight spaces) while the heavier fraction stays behind afterward to provide lubrication and surface protection after the light stuff evaporates.
Now that's great for things that need protection against corrosion but don't move. And it's also great for things like bearings and electric motors, which heat up and thin out the residual heavy fraction. But a gun - especially a slide-action, but also the bulled down the barrel - doesn't have time to heat it up and thin it out before a thick oil's interference would cause problems.
It's like putting your gun in cosmoline for storage. Great to prevent corrosion. But get it ALL out before you try to shoot it.
Note that I'm just repeating what I heard from a gunsmith, not speaking from a personal inside knowlege of the formulation or personal testing. (Thanks, but using it on my guns to see if they jam or blow up isn't my idea of fun.) Maybe he's repeating a false rumor. But I doubt he had any significant incentive to fib about it.
And I recall another gunsmith who would not let people fire "K-mart guns or gun-show ammunition" in his range, and was constantly ragging them. Story was that you couldn't trust even brand-name guns at discount stores, because the sporting goods depts would negotiate a low price on a boxcar-load and the manufacturers would, of course, sell them their seconds. (There's a lot of variation in gun manufacture.) Meanwhile, reloaders would sometimes goof and double-powder a batch of ammo - or think they might have - and sell them at a gun show rather than pull the bullets and try again.
My shooting buddy and I wondered whether he was overstating the case and downing the competition. Until one day we were at a state park shooting range and a guy came up with his brand-new K-mart gun and a batch of gun-show ammo. First round blew the rifle apart and cut up his hand something fierce. We compression-bandaged it with (new-from-the-bag) cleaning patches, his wife drove him off to the emergency room, and we never doubted that gunsmith again.
Now what if we (meaning the US) mistakenly elect government officials with very bad intentions?
Or what if some corrupt people rig an election? You don't even need the population to actually vote 'em in.
Not that it matters, though. As you point out: I will neglect specific examples in order to avoid Godwin's Law.
Note that the "maximum leader" of that specific example was WILDLY popular with the electorate - including some that he later sent to their deaths. The war-era and postwar propaganda makes him out to be very abrasive. But in fact he was a friendly and cuddly sort in most of his personal appearances. A real baby-kisser. Popular with the ladies. Charismatic. Animal rightist. Environmentalist. Body-beautiful. Rah-rah our oppressed country. Fight crime. Blah blah blah.
MOST dictators start out like that. And still seem like that to many of their constituents even at the height of their atrocities. Which is why the above unnamed leader won his last elected position by a considerable margin.
By the way:
Premise: "Those who cannot remember the past are condemned to repeat it."
Crollary: "Godwin's Law is very convenent for neofascists."
It WILL be back. It's ALWAYS trying to come back. When it comes, it will be called ANYTHING but the names it was called before. But it will be the same old thing.
Remember: "History doesn't repeat. But it does rhyme."
Meanwhile, the NRA is complaining that gun laws are getting progressively tougher. You don't seem to be in step with your compatriots there?
No.
Because on one hand CCW liberalization is occuring in SOME states. But on the other hand there are plenty of other draconian measures being proposed, and occasionally passed, in other states - and at the federal level - and in some of the same states that liberilized CCW laws - and sometimes as part of the provisions of the same "liberalized" CCW laws.
Some examples of that last: "mental health" exceptions that would open medical records to government scrutiny, ban anyone ever treated for depression (which makes them LESS likely to harm someone else, and would ban over half of women from carrying guns) or "post-traumatic stress" (which means anyone who went to an emergency-room after being attacked).
This is an ongoing battle, with the front advancing in some places and retreating in others. And it will probably continue forever. "The tree of liberty must be watered, from time to time..."
The only thing I can see that would be even CLOSE to a long-term victory would be if the Supreme Court finally heard a second-amendment case and decided for the "infringement means even touching the edges", striking down ALL gun regulations, including taxes (as some states exempt newspapers from sales tax) due to the use of tax laws as a backdoor regulation
But even that would just start the anti-gunners looking for a way around it - even if it meant amending the constitution.
I suspect that the parent was referring to the email potentially being spoofed, which may be considered ironic if the Security advisor appears to have resigned because of a security breach.
It would be even FUNNIER if the resignation was a forgery - but then he had to resign over it, making it a self-fulfilling forgery. B-)
WD-40 is very good for breaking things free and a number of other purposes. But it should NOT be used on firearms, especially semi-autos. With time it gums up and collects dirt significantly more than other alternatives. This may cause malfunction when you suddenly need it. (And, as with other thick oils, leaving a heavy coat in the barrel may cause an explosion by slowing the bullet on its way out.)
Use it to free a jam if you must. But clean it all out afterward and lubricate with an oil designed for the purpose. (A classic is Hoppes #9, which is a combination of a powder solvent and a light oil, letting you clean and lube in one pass.)
Actually that would be the right wing revolving door, as violent criminals have to be released to make room for non violent drug offenders who got stuck with right-wing minimum sentences. Which will only get worse if the right wing Orrin Hatch gets his way and judges loose even more disgression when sentencing.
I agree with you that they are definitely a BIG part of the problem.
I will take SOME issue with you on the matter of which "wing" is responsible. As I see it, historically they BOTH have pushed for such laws. But currently the only in-power voices for legalization seem to be Republicans.
But I'll be HAPPY to pat on the back any Democrat who proposes, or votes for, a legalization bill. Can you name one?
Eh, what? I don't see that proven at all. There are many different statistics and facts flying backward and forward on this one. The case is far from proven.
Saying it doesn't make it so.
The scholarly studies that survived peer review are the ones that showed reductions (much to the initial surprise of the researchers who conducted them). The others "flying backward and forward" have been pretty thoroughly debunked. And not just by the NRA. B-)
If you people in the US want to start going down that road, fair enough,
Actually we started out "down that road", but came "up" it - with steadily-increasing gun restrictions and increasing violence - for a long time. This started turning around in the late '90s.
On the state level "gun control" started right after the Civil War, with the jim-crow laws to disarm the freed slaves (and poor whites). Did you know that the full term is "Niggertown Saturday Night Special"? It comes from the debates on those same jim-crow laws, which banned affordable guns as suitable only "for Niggertown on a Saturday night".
but pray you don't end up in the same situation as South Africa. Car-jackers and thieves no longer bother asking questions first, they just shoot.
Note that, not long after the ending of Apartheid, the new integrated govenment of South Africa instituted significant gun restrictions - with the predictable results you mention.
Here in the US there was quite a bit of carjacking for a while, mostly in areas with draconian gun laws, after car alarms became common. And it virtually disappeared again in various states as CCW liberalization passed.
Florida's experience is particularly instructive: After they went to CCW-on-demand crime of many sorts dropped like a rock. Until a criminal gang got the bright idea of carjacking tourists in rented cars, especially those just leaving an airport. (People from out-of-state would be expected not to have guns, and federal regulations keep 'em from having guns in the airport.) Florida solved this by issuing CCWs to tourists and taking the distinctive plate off the rental cars. B-) Interestingly, even at the height of the much-trumpeted Florida tourist targeting, a tourist in Florida was at less risk of violent crime than one in the urban parts of California (where nobody but the Police Chief's cronies can carry a gun).
In Britain, guns are outlawed almost completely. Not even the police carry weapons.
Boy are YOU out-of-date. Violent crime in Britain has SKYROCKETED since the gun bans that fell out of that incident in Scotland. "Hot" burglaries (where armed crooks break into occupied houses), are the rule rather than the exception, and the police generally DO carry guns now (though unobtrusively) in many jurisdictions.
Meanwhile, over here in the US violent crime is dropping, as progressively more states liberalize CCW and more people in those states go through the necessary hoops to acquire the permit, training, and gun.
[people who talk about privacy-invasion leading to government abuse] end up sounding like the NRA. that's uncomfortable for millions who do not agree that everybody packing is a good idea.
Actually, what the NRA wants is for anyone who is law-abiding to be ABLE to pack if (s)he wants to. That way people going to/from a shooting sport or gun shop, or carrying to defend against a perceived threat, won't be inconvenienced.
A couple percent carrying concealed on any given day is enough to put a MAJOR dent in violent crime.
[the people uncomfortable with "everyone packing"] want some policing
The trouble with policing is that it ALWAYS gets used to oppress out-groups. Bearing arms is a RIGHT - guaranteed by the Constitution. Just like voting, the NRA members believe it should not be restricted to people of certain colors, members of certain political parties, or campaign contributors of the local police chief or sheriff.
ANY government requirement will be - indeed, has been - used to restrict law-abiding citizens in anti-gun jurisdictions. "Tests", for instance, become like the "literacy tests" once used to deny voting rights to blacks, or are administered rarely in difficult-to-reach places (denying everyone who doesn't have the money to take the time off and the connections to find out in time). "Mandatory training courses" are often made expensive and hard-to-take just like tests. When they're not, they're still an extra cost - and also reduce the public's expertese, because people who take them tend to STOP THERE, rather than getting more safety education on their own. "Registration" creates lists that are used for confiscation (as they were in New York City and California in recent history). "Police discression" means the local Sheriff/Police Chief gets to disarm ANYONE he wants.
All these factors tend to disarm the law-abiding population. Meanwhile the crooks have all the guns they want - stolen from citizens, from armories, bought from crooked cops, smuggled in disguised as harmless bales of marijuana or cocaine, or cranked out in a metal shop more rudimentary than that needed for a good brake job.
Now I disagree with the NRA's support of the "war on drugs"
As do many of the (millions of) members. Unfortunately, the current NRA administration is from the laws-and-orders faction.
and their concept that we generally need more imprisonment of all sorts of criminals (as opposed to basic economic changes that reduce the incentives for criminal behavior)
Actually that got started as a response to the left-wingers running a revolving-door justice system (so the violent offenders were constantly being dumped back on the streets), then using the resulting mayhem to call for more gun laws (allegedly to disarm these violent criminals, thus making the streets safe again). Of course since the gun laws just disarm the victims this leads to still more mayhem, in never-ending positive feedback.
The NRA "winning team"'s response was to call for keeping the violent offenders in the clink, in order to help take the pressure off the gunnies.
Of course now that it has been thoroughly proven that relaxing the gun laws so a small fraction of the population is carrying concealed at any given time REDUCES both crime and violence, and this fact is beginning to penetrate the general awareness, such a program is counter-productive.
I wondered about switching to Linux and how much that would *save* them.
I mentioned that [...] and she said that they discussed it many times, but they ran figures on how much money they spent/lost just switching from one *program* to another (training and help desk support), let alone to a whole new operating system [,,,]
Their concerns are genuine. But their experience has no doubt been largely with switching between one Microsoft- or Mainframe-based application and another. Things may have changed a lot.
It's a pity she's no longer with IBM. Since they're now spending billions on Linux support her department would have a well-funded in-house helper and upper-management buyin for an experiment the next time the issue came up. (And her department's management would get interdepartmental-cooperation brownie points for trying it, too.)
Such an experiment for IBM would be a benefit regardless of the outcome. If it failed, the Linux people could analyze why and help the open-source community fix it. If it succeeded they could trumpet it to the business world in their next press push. B-)
The main cost here would not be the licensing, but rather the training until the same level expertise is reached with the new system for the workstation user (lost man hours, actual cost of training etc.) and support costs.
Right (if the licencing costs are not REALLY high.) But the training is once and the support costs are ongoing. Support cost differences quickly dominate once you're over the hump.
[... assume] 100 users need a support staff of 3-5 people [...] The avg. college kid can probably work as an intern in a lot of these when it comes to M$ based solutions, but when you go off into the world of Unices, where people actually need to have a basic understanding of what is happening support costs (and the avg. wage of the staff) would skyrocket
I think you're off on using an intern for support. That misses the added costs incurred when he hits the problems he CAN'T handle correctly - both the added costs of worker/application/business-process downtime while he calls for more trained help and the added costs ditto.
But the BIG thing your analysis missed - which the TCO studies funded by others than Microsoft catch - is the effect of the higher reliability of open-source solutions. This reduces costs two ways:
First: Though you need people who know what they're doing, you need a MUCH SMALLER NUMBER of them, because they put in much less time per-machine.
Second: Because things don't fail as often, your business processes have LESS DOWNTIME. So you get back a LOT of productivity in those hundreds of workers who spend more of their time working and less of it waiting for the helpdesk.
That last factor is another component of why open-source has achieved penetration in servers first. Different functions have different costs of downtime. For a generic worker it varies a lot depending on your particular business and the workers function in it, while the costs accrue in one department and the benefits in another. For a server - especially a business-critical-function server - the costs of downtime are almost always very high, while the server is bought and administered by the same department that handles its maintenance, making all three components of its TCO visible to the same bosses.
Lex Talionis, the principle of an eye for an eye, is a morally bankrupt code of law we've been moving away from for the past few thousand years, thankfully.
Wrong. Lex Talionis was the principle that you take NO MORE than an eye for an eye - promulgated as an "improvement" in an era where the response to losing an eye (or a purse) might be to do in the alleged perpetrator and confiscate all his worldly goods.
It's morally bankrupt, all right. But only to the extent that if the thief only loses what he stole, and has a nonzero chance of getting away with it, theft remains a profitmaking enterprise despite full enforcement of the law. So it becomes an endorsement of theft as a lifestyle. This is why there are "puntitive damages" - extra penalties to punish the perpetrator (thus making continued misbehavior a losing proposition even with imperfect law enforcement).
None of which applies here. Applying "Lex Talionis" to the spammer would mean spamming him, rather than seeking compensatory and puntitive damages.
===
Which is what they did, isn't it? B-)
===
Lex Talionis also recognizes a moral principal of equivalency, to wit: In an egalitarian society, regardless of what actions you think are fair, you have NO moral gripe if someone does to YOU what YOU did to them. If it was wrong for them to do in retaliation, it was AT LEAST as wrong for YOU to do without provocation.
===
I note, by the way, that your posting is IDENTICAL to one you made severaltimespreviously - including in the slashdot article credited with inspring the USPS DDoS attack in the first place. (And that last one I cited was under your own slashdot ID of Chuck Flynn.) Given that, I felt free to repeat, almost verbatim, my response to your most recent previous missive.
The posts that recieve your canned response seem to be any suggestion about spamming the spammers. You wouldn't happen to be a spammer, would you?
Quality Control will be a nightmare...99.99% accuracy will result in 100 errors per million units produced. That is 100 errors straight off the production line. Imagine what will be happening to the things as they age.
The upside is that nanomachines - defined as mechanical devices whose working parts are single molecules - are subject to breakage but NOT wear. Unlike macro devices their individual parts are not lump-of-clay-like collections of smaller parts, that can work well in many configurations. Nanomachine parts are either correctly connected or not.
Thus they are more like digital systems (where the answer is either right or wrong) than analog (where the answer may have varying degrees of error).
This all-or-nothing nature, in combination with the reuse of single components for many functions, makes it possible to design in self-diagnostic tools. The processor and storage portion of the device can use error-checking codes to detect corruption (reducing the incidence of undetected corruption to arbitrarily low levels), while the mechanical parts can be included in diagnostic routines that they will fail if even a single atom is out of place or a single component missing, doubled, or out-of-position.
With components being churned out by the billion, stop-forever-on-any-failure, march-to-shipping (or start executing the redundancy-checked main program) -on-pass is adequate testing to assure that only good machines get into use. "Broken" machines can be scavenged - down to the atomic level - to clean up the "manufacturing floor" without perpetuating their breakage in devices involved in their salvage or constructed from their remains.
As for aging: Redundancy checks (and halt-on-failure) can trivially reduce the likelyhood of an undetected software failure to arbitrarily low levels, just as they did in the original manufacturing process. Similarly, periodic rerunning of the hardware diagnostics can do the same for the physical components.
(There's a theory that the observed tendency of each foetus to go through the evolutionary stages of the organism's history during its development serves as a similar "manufacturing QA test" that the important genetic systems to be used later in life are correctly in place - with the foetus dying early if a major system is broken or missing. It works well. But designed systems can do better.)
Wasn't this suppose to be a joke question? A play on the breakfast cereal character Captain Crunch? Yet pumped up as interesting..
No. It's a reference to why John's phreaker nick was Captain Crunch, back in the middle of the Twentieth Century.
In those days the long distance system used in-band signaling tones to connect, disconnect, and dial calls. The base system used a 2600 Hz tone to do the rough equivalent of "on-hook". (That's why 2600 magazine is named that.) The full-blown system also had a set of dual-tones, similar to touch-tones, to "dial" the call - but on some trunks (typically those going to legacy dial-only exchanges) you also dialed the call by switching the 2600 Hz tone on and off like a pulse dialer.
Phone Phreaks needed a tone generator to do the dual-tone system. But you could whistle the 2600 "disconnect" tone (especially if you had perfect pitch). If you were REALLY good you could also whistle pluses of it to dial calls. But that was tough. Something over 5 pulses per second or they separate into two digits. Get every digit right or you dial the wrong number.
Then Captain Crunch cereal came out with a prize inside - a plastic whistle. It JUST HAPPENED to be 2600 Hz. Oops! With the whistle it was REALLY EASY to "blow off" calls and/or to dial calls on the legacy-exchange trunks. John Draper noticed this, made heavy use of it, became famous in phreaker circles for it, and eventually used Captain Crunch as his phone phreak nickname.
So his first Captain Crunch plastic whistle is a real historical artifact. (And probably sitting in an evidence locker somewhere if he didn't lose it long ago.)
You can find 512 MB flash cards now for about $100 and the price is dropping by at least a factor of two every 10 months. [...] I don't see why anyone would buy this.
Flash is starting to hit some physical limits and may fall off Moore's Law real soon now, unless they come up with a new storage mechanism. Current flash depends on stored charge behind an insulator, and you can't scale that down much more because the insulator is already thin enough that the electrons are on the edge of tunneling through. (That's how you write it.)
Mag disk, on the other hand, has a long way to go before it hits a limit. And the size of the disk doesn't have a lot to do with its capacity, since reduced size also means reduced length of parts to be affected by tolerances and to sonic-delay the part of the feedback loop from actuator to head, so tolerances cancel out and the limit is the size of the bit on the surface - currently limited by head technology rather than physics of magnetization.
I don't really understand why this thing ONLY has 1.5G, unless they didn't want to push the envelope on the first release. Expect that to climb rapidly if there's any pressure from flash.
So basically its a propriatory interface. Its cool don't get me wrong but I don't think IBM will be scared just yet.
That depends on the interface, doesn't it? If it's dog-simple to support on the far end it might take off big time. If they provide a small macro for designers to use in FPGAs or ASICs, standards aren't a major issue. Ditto if it presents itself as an internet-like device you can get to through a stock serial port and a minimal stock stack.
Looks like five wires. Five? Power, ground, three left over. Clock, TxD, Rxd? Bidirectional balanced serial bus and a reset/shutdown signal? Motor power, logic power, ground, bidir bus? Power, ground, balanced bus, EMI ground? (Maybe the flexy-board is double sided and it's ten wires?)
I want to see a description of this "simple and original" interface.
The parent post makes the mistake of identifying bubble formation with the cavitation damage, where as you point out, it is the bubble collapse that is the dangerous part.
Bubbles forming on a hot spot in a liquid that is significantly below its boiling point collapse in place. The expansion gives them a large cooling surface an allows the vapor to suddenly cool below the boiling point and recondense. It isn't until the liquid is superheating near the bubble formation site that the bubble continues to expand and breaks off rather than collapsing.
Watch a pot of water coming to a boil some time. It goes through a phase where the bubbles flicker in place, before warming to the point where the bubbles break off and float away.
Congress shall make no law [...] abridging the freedom of speech, or of the press; [...]
Sounds to me like "you can say what you want, when you want, and no consequences" to me.
What you want, yes. When you want, yes. No consequences, no.
The amendment has been interpreted to mean that the congress can't stop you ahead of time, but can set up rules for punishing you after the fact if your speech meets certain criteria. (Like harming others, soliciting crimes, or otherwise interfering with a "compelling state interest".)
While I'm with you on this one (the GOVERNMENT shouldn't be setting up any content-based penalties for speech, before or after the act), the Supreme Court says otherwise. And there's no appeal beyond the supreme court - which is why it gets to rule on the constitutionality of laws and have the rules stick.
(Oh, well. They say two out of three ain't bad...)
Slashdot Mirror
I was always under the impression that patents were for inventions and not ideas
That's right.
[...] (minus this crap about software patents).
Software is only patentable when it is an "invention". For a long time it wasn't patentable at all. If I recall correctly, the logjam broke when somebody snuck one through by describing a hardware device to do the same computation, then griped about it in court. And another patented an invention that included a software-driven component.
IMHO a piece of software is simultaneously a "mathematical algorithm" and a "work of composition" but not a "performance" (even if it DOES prescribe a machine's actions). So it should be unpatentable, the source code should be copyrightable (though the current near-infinite copyrights are nuts), and the copyright should carry over to the object as a "derived work". And it should be subject to reverse-engineering of how it works, fair use of tiny snippets of code, but not verbatim plagarism of the whole or large chunks of code.
What if you write some vague description of a computer program or some device, but can't figure out how to make it work with all the details in it? That's where the problem comes in.
Indeed. But part of the requirements that the patent examiner is checking is that the description is adequate for "someone expert in the art" to "perform the invention". Sometimes it's complicated enough, or enough beyond the examiner's expertese, that the examiner (who has limited time) lets it go through even if it really is broken or incompelete. (And that's how the occasional perpetual motion machine still gets patented.) But
They should require working models of the patented products (in the case of software: working source code if there really must be software patents)) so that you're sure that the person patenting the invention has done the real work of inventing something.
Why? They DID that. And they got BURIED in working models.
What do you do when making the first one costs billiions? Say: An orbital skyhook design? Or how about the "whip" satelite launcher? (Trail the satelite behind a 747 on a long cord, do a loop and "crack the whip", releasing the satelite when it's headed beyond the atmosphere and the 747 is flying a couple miles lower.) Are you saying deliver a skyhook or a 747 with a whip to the patent office or forego the patent?
What about a gene-engineered bacterium that kills colon cancer? A "working model" would include the patients of the human trials.
And the time to patent a lot of stuff is when it's still in the design stages - when you know WHAT you're doing but haven't had a hundred-man team dot the Is and cross the Ts of the implementation. (And three of 'em quit and go work for the competition.) "Working source code" is right out in such a situation.
One takes a look at all the patents given in the US that don't and can NEVER be produced because they are physically impossible.
It's not the job of the patent office to determine whether the invention actually works, or can even be constructed.
It's the job of the patent office to certify that, within the results of a reasonable search, what you registered is not already registered.
This certification is a time-limited "license to sue" anyone else who does the same thing the same way - and a presumption that you have precedence. (i.e. it's up to the other guy to prove that his thing is different, or that he had it before you did.)
A patent on something impossible is just a number used up in an aleph-null namespace, a few documents on file in a bureaucracy, and some manhours that were paid for by the guy who patented the impossible thing. In theory, nobody is going to infringe a patent-on-the-impossible, so no court case (unless the fruitcake starts suing people randomly, of course).
Single exception: Perpetual motion machines. The patent office was SO clogged with perpetual motion applications in the steam-engine era that they bogged down. So they still require a working model. (As, back when they started up, they did for everthing, before the office became clogged with working models. Which is where the Smithsonian got a lot of their exhibits. B-) )
Routinely bouncing putative perpetual-motion devices sometimes is a problem. For instance, an inventor came up with a VERY efficient still: Very tall, so the top is at a near-vacuum due to the weight of the liquids. Counter-current heat excanger transfers heat from the the condensate and brine to the feedstock, so most of the heat is recycled and the product and waste come out cool. Base-of-couumn pressure difference between the feedstock and the outputs, to create the temperature difference across the heat exchanger.
Patent office bounced it for being a perpetual-motion mcahine. Inventor got his patent after proving that you still had to energy to provide the heat-of-solution of the impurities and to replace losses in the (necessarily) less-than-perfect heat exchanger and through the insulation between the column and the environment. It was VERY GOOD - because the input heat got used many times rather than once - but it wasn't free.
It's nice to see that Linus has his priorities in the right place. Too bad others can't follow his example. *cough* SCO *cough*
Let's not misinterpret SCO's actions.
They're claiming that IBM took some of their licensed software and, in violation of the license, purported to sublicense rights they didn't have under the GPL.
Unlike the original UNIX (System N), whose proprietary status may be problematic, SVRn was carefully re-implemented under well-defined IP law. (In fact, some believe that was the whole POINT of SVRn.) It is NOT public unless its owner MAKES it public. Right now its owner is SCO.
If there really is proprietary code circulating with GPL copyleft attached, it needs to be identified, purged, and replaced - or otherwise made free.
For instance: IBM might counter-sue and the court might find that SCO had incorporated GPLed code with functionality of comparable value, and as part of the decision make the leaked SCO code public. (Courts DO like to upset the status-quo as little as possible if it can be done equitably.)
we all know you should wipe down after cleaning with whatever material you choose to apply. A spider web in the barrel of a shotgun can be lethal. There is always a need for care and common sense when dealing with any weapon.
Hear, hear.
I also note that the WD-40 faq you quoted does recommend it for guns - for protection from corrosion - and claims it won't harm bluing. But it doesn't recommend it for lubrication. The same site DOES recommend another of their products - 3-in-1 white lithium grease - for lubricating sliding parts on weather/dirt-exposed machinery - specifically including guns. (They also recommend their 3-in-1 silicone spray for protecting gun finishd surfaces from corrosion.)
And with Microsoft's latest effort to try to make their OS's as "secure" as possible, shouldn't all these people picking on opensource be targeting Microsoft as well, since they are now SECURE?
You mean the one where they give copies of the source code to the information warfare departments of all the major powers (including the US's former enemies - Russia and China), but still keep it a "secret" from the public-sector crypto and software scholars?
B-)
WD-40 is _not_ a thick oil. It has a carrier that evaporates, and is one of the thinnest, which is why it is sold as an aerosol. You're clearly confusing it with some other product.
My understanding is that WD-40 is a MIX of very light and very heavy fractions. The light fractions (which are the bulk of it) aid penetration and freeing of stuck parts (by disolving older tar-like stuff in the tight spaces) while the heavier fraction stays behind afterward to provide lubrication and surface protection after the light stuff evaporates.
Now that's great for things that need protection against corrosion but don't move. And it's also great for things like bearings and electric motors, which heat up and thin out the residual heavy fraction. But a gun - especially a slide-action, but also the bulled down the barrel - doesn't have time to heat it up and thin it out before a thick oil's interference would cause problems.
It's like putting your gun in cosmoline for storage. Great to prevent corrosion. But get it ALL out before you try to shoot it.
Note that I'm just repeating what I heard from a gunsmith, not speaking from a personal inside knowlege of the formulation or personal testing. (Thanks, but using it on my guns to see if they jam or blow up isn't my idea of fun.) Maybe he's repeating a false rumor. But I doubt he had any significant incentive to fib about it.
And I recall another gunsmith who would not let people fire "K-mart guns or gun-show ammunition" in his range, and was constantly ragging them. Story was that you couldn't trust even brand-name guns at discount stores, because the sporting goods depts would negotiate a low price on a boxcar-load and the manufacturers would, of course, sell them their seconds. (There's a lot of variation in gun manufacture.) Meanwhile, reloaders would sometimes goof and double-powder a batch of ammo - or think they might have - and sell them at a gun show rather than pull the bullets and try again.
My shooting buddy and I wondered whether he was overstating the case and downing the competition. Until one day we were at a state park shooting range and a guy came up with his brand-new K-mart gun and a batch of gun-show ammo. First round blew the rifle apart and cut up his hand something fierce. We compression-bandaged it with (new-from-the-bag) cleaning patches, his wife drove him off to the emergency room, and we never doubted that gunsmith again.
Did you ever find WD-40 at a gun shop?
Now what if we (meaning the US) mistakenly elect government officials with very bad intentions?
Or what if some corrupt people rig an election? You don't even need the population to actually vote 'em in.
Not that it matters, though. As you point out: I will neglect specific examples in order to avoid Godwin's Law.
Note that the "maximum leader" of that specific example was WILDLY popular with the electorate - including some that he later sent to their deaths. The war-era and postwar propaganda makes him out to be very abrasive. But in fact he was a friendly and cuddly sort in most of his personal appearances. A real baby-kisser. Popular with the ladies. Charismatic. Animal rightist. Environmentalist. Body-beautiful. Rah-rah our oppressed country. Fight crime. Blah blah blah.
MOST dictators start out like that. And still seem like that to many of their constituents even at the height of their atrocities. Which is why the above unnamed leader won his last elected position by a considerable margin.
By the way:
Premise: "Those who cannot remember the past are condemned to repeat it."
Crollary: "Godwin's Law is very convenent for neofascists."
It WILL be back. It's ALWAYS trying to come back. When it comes, it will be called ANYTHING but the names it was called before. But it will be the same old thing.
Remember: "History doesn't repeat. But it does rhyme."
Meanwhile, the NRA is complaining that gun laws are getting progressively tougher. You don't seem to be in step with your compatriots there?
..."
No.
Because on one hand CCW liberalization is occuring in SOME states. But on the other hand there are plenty of other draconian measures being proposed, and occasionally passed, in other states - and at the federal level - and in some of the same states that liberilized CCW laws - and sometimes as part of the provisions of the same "liberalized" CCW laws.
Some examples of that last: "mental health" exceptions that would open medical records to government scrutiny, ban anyone ever treated for depression (which makes them LESS likely to harm someone else, and would ban over half of women from carrying guns) or "post-traumatic stress" (which means anyone who went to an emergency-room after being attacked).
This is an ongoing battle, with the front advancing in some places and retreating in others. And it will probably continue forever. "The tree of liberty must be watered, from time to time
The only thing I can see that would be even CLOSE to a long-term victory would be if the Supreme Court finally heard a second-amendment case and decided for the "infringement means even touching the edges", striking down ALL gun regulations, including taxes (as some states exempt newspapers from sales tax) due to the use of tax laws as a backdoor regulation
But even that would just start the anti-gunners looking for a way around it - even if it meant amending the constitution.
I suspect that the parent was referring to the email potentially being spoofed, which may be considered ironic if the Security advisor appears to have resigned because of a security breach.
It would be even FUNNIER if the resignation was a forgery - but then he had to resign over it, making it a self-fulfilling forgery. B-)
WD-40 is very good for breaking things free and a number of other purposes. But it should NOT be used on firearms, especially semi-autos. With time it gums up and collects dirt significantly more than other alternatives. This may cause malfunction when you suddenly need it. (And, as with other thick oils, leaving a heavy coat in the barrel may cause an explosion by slowing the bullet on its way out.)
Use it to free a jam if you must. But clean it all out afterward and lubricate with an oil designed for the purpose. (A classic is Hoppes #9, which is a combination of a powder solvent and a light oil, letting you clean and lube in one pass.)
Actually that would be the right wing revolving door, as violent criminals have to be released to make room for non violent drug offenders who got stuck with right-wing minimum sentences. Which will only get worse if the right wing Orrin Hatch gets his way and judges loose even more disgression when sentencing.
I agree with you that they are definitely a BIG part of the problem.
I will take SOME issue with you on the matter of which "wing" is responsible. As I see it, historically they BOTH have pushed for such laws. But currently the only in-power voices for legalization seem to be Republicans.
But I'll be HAPPY to pat on the back any Democrat who proposes, or votes for, a legalization bill. Can you name one?
Eh, what? I don't see that proven at all. There are many different statistics and facts flying backward and forward on this one. The case is far from proven.
Saying it doesn't make it so.
The scholarly studies that survived peer review are the ones that showed reductions (much to the initial surprise of the researchers who conducted them). The others "flying backward and forward" have been pretty thoroughly debunked. And not just by the NRA. B-)
If you people in the US want to start going down that road, fair enough,
Actually we started out "down that road", but came "up" it - with steadily-increasing gun restrictions and increasing violence - for a long time. This started turning around in the late '90s.
On the state level "gun control" started right after the Civil War, with the jim-crow laws to disarm the freed slaves (and poor whites). Did you know that the full term is "Niggertown Saturday Night Special"? It comes from the debates on those same jim-crow laws, which banned affordable guns as suitable only "for Niggertown on a Saturday night".
but pray you don't end up in the same situation as South Africa. Car-jackers and thieves no longer bother asking questions first, they just shoot.
Note that, not long after the ending of Apartheid, the new integrated govenment of South Africa instituted significant gun restrictions - with the predictable results you mention.
Here in the US there was quite a bit of carjacking for a while, mostly in areas with draconian gun laws, after car alarms became common. And it virtually disappeared again in various states as CCW liberalization passed.
Florida's experience is particularly instructive: After they went to CCW-on-demand crime of many sorts dropped like a rock. Until a criminal gang got the bright idea of carjacking tourists in rented cars, especially those just leaving an airport. (People from out-of-state would be expected not to have guns, and federal regulations keep 'em from having guns in the airport.) Florida solved this by issuing CCWs to tourists and taking the distinctive plate off the rental cars. B-) Interestingly, even at the height of the much-trumpeted Florida tourist targeting, a tourist in Florida was at less risk of violent crime than one in the urban parts of California (where nobody but the Police Chief's cronies can carry a gun).
In Britain, guns are outlawed almost completely. Not even the police carry weapons.
Boy are YOU out-of-date. Violent crime in Britain has SKYROCKETED since the gun bans that fell out of that incident in Scotland. "Hot" burglaries (where armed crooks break into occupied houses), are the rule rather than the exception, and the police generally DO carry guns now (though unobtrusively) in many jurisdictions.
Meanwhile, over here in the US violent crime is dropping, as progressively more states liberalize CCW and more people in those states go through the necessary hoops to acquire the permit, training, and gun.
[people who talk about privacy-invasion leading to government abuse] end up sounding like the NRA. that's uncomfortable for millions who do not agree that everybody packing is a good idea.
Actually, what the NRA wants is for anyone who is law-abiding to be ABLE to pack if (s)he wants to. That way people going to/from a shooting sport or gun shop, or carrying to defend against a perceived threat, won't be inconvenienced.
A couple percent carrying concealed on any given day is enough to put a MAJOR dent in violent crime.
[the people uncomfortable with "everyone packing"] want some policing
The trouble with policing is that it ALWAYS gets used to oppress out-groups. Bearing arms is a RIGHT - guaranteed by the Constitution. Just like voting, the NRA members believe it should not be restricted to people of certain colors, members of certain political parties, or campaign contributors of the local police chief or sheriff.
ANY government requirement will be - indeed, has been - used to restrict law-abiding citizens in anti-gun jurisdictions. "Tests", for instance, become like the "literacy tests" once used to deny voting rights to blacks, or are administered rarely in difficult-to-reach places (denying everyone who doesn't have the money to take the time off and the connections to find out in time). "Mandatory training courses" are often made expensive and hard-to-take just like tests. When they're not, they're still an extra cost - and also reduce the public's expertese, because people who take them tend to STOP THERE, rather than getting more safety education on their own. "Registration" creates lists that are used for confiscation (as they were in New York City and California in recent history). "Police discression" means the local Sheriff/Police Chief gets to disarm ANYONE he wants.
All these factors tend to disarm the law-abiding population. Meanwhile the crooks have all the guns they want - stolen from citizens, from armories, bought from crooked cops, smuggled in disguised as harmless bales of marijuana or cocaine, or cranked out in a metal shop more rudimentary than that needed for a good brake job.
Now I disagree with the NRA's support of the "war on drugs"
As do many of the (millions of) members. Unfortunately, the current NRA administration is from the laws-and-orders faction.
and their concept that we generally need more imprisonment of all sorts of criminals (as opposed to basic economic changes that reduce the incentives for criminal behavior)
Actually that got started as a response to the left-wingers running a revolving-door justice system (so the violent offenders were constantly being dumped back on the streets), then using the resulting mayhem to call for more gun laws (allegedly to disarm these violent criminals, thus making the streets safe again). Of course since the gun laws just disarm the victims this leads to still more mayhem, in never-ending positive feedback.
The NRA "winning team"'s response was to call for keeping the violent offenders in the clink, in order to help take the pressure off the gunnies.
Of course now that it has been thoroughly proven that relaxing the gun laws so a small fraction of the population is carrying concealed at any given time REDUCES both crime and violence, and this fact is beginning to penetrate the general awareness, such a program is counter-productive.
I wondered about switching to Linux and how much that would *save* them.
I mentioned that [...] and she said that they discussed it many times, but they ran figures on how much money they spent/lost just switching from one *program* to another (training and help desk support), let alone to a whole new operating system [,,,]
Their concerns are genuine. But their experience has no doubt been largely with switching between one Microsoft- or Mainframe-based application and another. Things may have changed a lot.
It's a pity she's no longer with IBM. Since they're now spending billions on Linux support her department would have a well-funded in-house helper and upper-management buyin for an experiment the next time the issue came up. (And her department's management would get interdepartmental-cooperation brownie points for trying it, too.)
Such an experiment for IBM would be a benefit regardless of the outcome. If it failed, the Linux people could analyze why and help the open-source community fix it. If it succeeded they could trumpet it to the business world in their next press push. B-)
The main cost here would not be the licensing, but rather the training until the same level expertise is reached with the new system for the workstation user (lost man hours, actual cost of training etc.) and support costs.
Right (if the licencing costs are not REALLY high.) But the training is once and the support costs are ongoing. Support cost differences quickly dominate once you're over the hump.
[... assume] 100 users need a support staff of 3-5 people [...] The avg. college kid can probably work as an intern in a lot of these when it comes to M$ based solutions, but when you go off into the world of Unices, where people actually need to have a basic understanding of what is happening support costs (and the avg. wage of the staff) would skyrocket
I think you're off on using an intern for support. That misses the added costs incurred when he hits the problems he CAN'T handle correctly - both the added costs of worker/application/business-process downtime while he calls for more trained help and the added costs ditto.
But the BIG thing your analysis missed - which the TCO studies funded by others than Microsoft catch - is the effect of the higher reliability of open-source solutions. This reduces costs two ways:
First: Though you need people who know what they're doing, you need a MUCH SMALLER NUMBER of them, because they put in much less time per-machine.
Second: Because things don't fail as often, your business processes have LESS DOWNTIME. So you get back a LOT of productivity in those hundreds of workers who spend more of their time working and less of it waiting for the helpdesk.
That last factor is another component of why open-source has achieved penetration in servers first. Different functions have different costs of downtime. For a generic worker it varies a lot depending on your particular business and the workers function in it, while the costs accrue in one department and the benefits in another. For a server - especially a business-critical-function server - the costs of downtime are almost always very high, while the server is bought and administered by the same department that handles its maintenance, making all three components of its TCO visible to the same bosses.
Lex Talionis, the principle of an eye for an eye, is a morally bankrupt code of law we've been moving away from for the past few thousand years, thankfully.
Wrong. Lex Talionis was the principle that you take NO MORE than an eye for an eye - promulgated as an "improvement" in an era where the response to losing an eye (or a purse) might be to do in the alleged perpetrator and confiscate all his worldly goods.
It's morally bankrupt, all right. But only to the extent that if the thief only loses what he stole, and has a nonzero chance of getting away with it, theft remains a profitmaking enterprise despite full enforcement of the law. So it becomes an endorsement of theft as a lifestyle. This is why there are "puntitive damages" - extra penalties to punish the perpetrator (thus making continued misbehavior a losing proposition even with imperfect law enforcement).
None of which applies here. Applying "Lex Talionis" to the spammer would mean spamming him, rather than seeking compensatory and puntitive damages.
===
Which is what they did, isn't it? B-)
===
Lex Talionis also recognizes a moral principal of equivalency, to wit: In an egalitarian society, regardless of what actions you think are fair, you have NO moral gripe if someone does to YOU what YOU did to them. If it was wrong for them to do in retaliation, it was AT LEAST as wrong for YOU to do without provocation.
===
I note, by the way, that your posting is IDENTICAL to one you made several times previously - including in the slashdot article credited with inspring the USPS DDoS attack in the first place. (And that last one I cited was under your own slashdot ID of Chuck Flynn.) Given that, I felt free to repeat, almost verbatim, my response to your most recent previous missive.
The posts that recieve your canned response seem to be any suggestion about spamming the spammers. You wouldn't happen to be a spammer, would you?
Quality Control will be a nightmare...99.99% accuracy will result in 100 errors per million units produced. That is 100 errors straight off the production line. Imagine what will be happening to the things as they age.
The upside is that nanomachines - defined as mechanical devices whose working parts are single molecules - are subject to breakage but NOT wear. Unlike macro devices their individual parts are not lump-of-clay-like collections of smaller parts, that can work well in many configurations. Nanomachine parts are either correctly connected or not.
Thus they are more like digital systems (where the answer is either right or wrong) than analog (where the answer may have varying degrees of error).
This all-or-nothing nature, in combination with the reuse of single components for many functions, makes it possible to design in self-diagnostic tools. The processor and storage portion of the device can use error-checking codes to detect corruption (reducing the incidence of undetected corruption to arbitrarily low levels), while the mechanical parts can be included in diagnostic routines that they will fail if even a single atom is out of place or a single component missing, doubled, or out-of-position.
With components being churned out by the billion, stop-forever-on-any-failure, march-to-shipping (or start executing the redundancy-checked main program) -on-pass is adequate testing to assure that only good machines get into use. "Broken" machines can be scavenged - down to the atomic level - to clean up the "manufacturing floor" without perpetuating their breakage in devices involved in their salvage or constructed from their remains.
As for aging: Redundancy checks (and halt-on-failure) can trivially reduce the likelyhood of an undetected software failure to arbitrarily low levels, just as they did in the original manufacturing process. Similarly, periodic rerunning of the hardware diagnostics can do the same for the physical components.
(There's a theory that the observed tendency of each foetus to go through the evolutionary stages of the organism's history during its development serves as a similar "manufacturing QA test" that the important genetic systems to be used later in life are correctly in place - with the foetus dying early if a major system is broken or missing. It works well. But designed systems can do better.)
Actually Crunch didn't discover the whistle, and I may have had a couple other items a bit off.
Here's the story in his words (via google's cache).
Wasn't this suppose to be a joke question? A play on the breakfast cereal character Captain Crunch? Yet pumped up as interesting..
No. It's a reference to why John's phreaker nick was Captain Crunch, back in the middle of the Twentieth Century.
In those days the long distance system used in-band signaling tones to connect, disconnect, and dial calls. The base system used a 2600 Hz tone to do the rough equivalent of "on-hook". (That's why 2600 magazine is named that.) The full-blown system also had a set of dual-tones, similar to touch-tones, to "dial" the call - but on some trunks (typically those going to legacy dial-only exchanges) you also dialed the call by switching the 2600 Hz tone on and off like a pulse dialer.
Phone Phreaks needed a tone generator to do the dual-tone system. But you could whistle the 2600 "disconnect" tone (especially if you had perfect pitch). If you were REALLY good you could also whistle pluses of it to dial calls. But that was tough. Something over 5 pulses per second or they separate into two digits. Get every digit right or you dial the wrong number.
Then Captain Crunch cereal came out with a prize inside - a plastic whistle. It JUST HAPPENED to be 2600 Hz. Oops! With the whistle it was REALLY EASY to "blow off" calls and/or to dial calls on the legacy-exchange trunks. John Draper noticed this, made heavy use of it, became famous in phreaker circles for it, and eventually used Captain Crunch as his phone phreak nickname.
So his first Captain Crunch plastic whistle is a real historical artifact. (And probably sitting in an evidence locker somewhere if he didn't lose it long ago.)
You can find 512 MB flash cards now for about $100 and the price is dropping by at least a factor of two every 10 months. [...] I don't see why anyone would buy this.
Flash is starting to hit some physical limits and may fall off Moore's Law real soon now, unless they come up with a new storage mechanism. Current flash depends on stored charge behind an insulator, and you can't scale that down much more because the insulator is already thin enough that the electrons are on the edge of tunneling through. (That's how you write it.)
Mag disk, on the other hand, has a long way to go before it hits a limit. And the size of the disk doesn't have a lot to do with its capacity, since reduced size also means reduced length of parts to be affected by tolerances and to sonic-delay the part of the feedback loop from actuator to head, so tolerances cancel out and the limit is the size of the bit on the surface - currently limited by head technology rather than physics of magnetization.
I don't really understand why this thing ONLY has 1.5G, unless they didn't want to push the envelope on the first release. Expect that to climb rapidly if there's any pressure from flash.
So basically its a propriatory interface. Its cool don't get me wrong but I don't think IBM will be scared just yet.
That depends on the interface, doesn't it? If it's dog-simple to support on the far end it might take off big time. If they provide a small macro for designers to use in FPGAs or ASICs, standards aren't a major issue. Ditto if it presents itself as an internet-like device you can get to through a stock serial port and a minimal stock stack.
Looks like five wires. Five? Power, ground, three left over. Clock, TxD, Rxd? Bidirectional balanced serial bus and a reset/shutdown signal? Motor power, logic power, ground, bidir bus? Power, ground, balanced bus, EMI ground? (Maybe the flexy-board is double sided and it's ten wires?)
I want to see a description of this "simple and original" interface.
The parent post makes the mistake of identifying bubble formation with the cavitation damage, where as you point out, it is the bubble collapse that is the dangerous part.
Bubbles forming on a hot spot in a liquid that is significantly below its boiling point collapse in place. The expansion gives them a large cooling surface an allows the vapor to suddenly cool below the boiling point and recondense. It isn't until the liquid is superheating near the bubble formation site that the bubble continues to expand and breaks off rather than collapsing.
Watch a pot of water coming to a boil some time. It goes through a phase where the bubbles flicker in place, before warming to the point where the bubbles break off and float away.
Congress shall make no law [...] abridging the freedom of speech, or of the press; [...]
Sounds to me like "you can say what you want, when you want, and no consequences" to me.
What you want, yes.
When you want, yes.
No consequences, no.
The amendment has been interpreted to mean that the congress can't stop you ahead of time, but can set up rules for punishing you after the fact if your speech meets certain criteria. (Like harming others, soliciting crimes, or otherwise interfering with a "compelling state interest".)
While I'm with you on this one (the GOVERNMENT shouldn't be setting up any content-based penalties for speech, before or after the act), the Supreme Court says otherwise. And there's no appeal beyond the supreme court - which is why it gets to rule on the constitutionality of laws and have the rules stick.
(Oh, well. They say two out of three ain't bad...)