^^^ argh. This is what happens when you post replies to Slashdot while eating breakfast, before the caffeine kicks in.
I went back and looked at my code from a few months ago. I don't have time to re-study it now to figure out what went wrong (I'm already late for work), but from what I remember, I either couldn't get it to do Cipher.getInstance() without throwing an exception, I had problems because Android only includes a partial implementation of Bouncycastle, and trying to drop the entire jarfile caused classloading problems, or maybe it just didn't work. I just remember giving up at some point and deciding it was a lost cause.
Here's my source, if anyone wants to play with it:
Cipher rsa = null;
byte[] password = (new String("passwordtest")).getBytes();
This is a more or less verbatim repeat of what I said to someone else, but it merits repeating here because your post addresses my response's topic head-on: the fundamental problem with that approach is that RSA and AES both include an element of random variation in the encrypted text. In other words, given a plaintext password and a public key, you can't assume that the encrypted output you get THIS time will be identical to the encrypted output you'll get NEXT time (or from a different implementation of the encryption algorithm). The encrypted output will decrypt with the private key just fine, but basically this means you can't treat the encrypted output like a reversible hash that just happens to take a lot longer to compute.
From what I remember from my research into this specific issue a few months ago, it's technically possible to write an implementation of RSA's encryption algorithm that introduces exactly zero bytes of random salt into the encrypted output, but I wasn't able to make it work using Bouncycastle, and got the impression that you can't do it with Microsoft's implementations or any other commonly-used implementation for Java, either.
AFAIK, there's no way at all to encrypt using AES without random entropy. It's an inherent part of the algorithm. In RSA's case, it's technically possible because it was tacked on as an afterthought, and things encrypted using the original algorithm end up looking like something encrypted using the new algorithm with zero bytes of random salt.
I'm not sure about other asymmetric algorithms. I remember hearing a lot about Elgamal back when I was in college, but ever since RSA's patent issues went away and AES became officially blessed by the US federal government, pretty much everything besides RSA and AES seem to have fallen by the wayside.
I had a similar idea a few months ago. I thought about using asymmetric encryption, and treating the values encrypted with the public key like hashes -- storing the encrypted values in the database, and keeping the private key safe and used only for offline decryption if necessary to fix a bug or something. Then I discovered the big problem -- AES and RSA include an element of randomness in their output. In other words, given the same input value and public key, you will NOT necessarily get the same encrypted output every time. Whatever you get WILL decrypt properly with the private key, but you can't assume that the encrypted output of a given input + public key you got this time will be identical to the encrypted output you'll get next time.
From what I remember, eliminating that randomness is impossible with AES (it's an inherent part of the algorithm). It might be possible with RSA *if* you had your own implementation of the encryption algorithm, but all the common implementations used by normal people (Microsoft, Bouncycastle, IBM, etc) do it. I'm not sure about Elgamal, or other algorithms that have fallen out of favor since RSA's patent expired and AES became the official standard of the US Federal Government.
I'm going to the launch, too. Realistically, how close do you have to BE in order to really experience it (hear the engines, be able to see it without binoculars, or at least without expensive binoculars, etc)? The closest I've ever come to seeing a shuttle launch in person was back around '94, when there was a night launch on a rare cloudless night, and I was able to stand on the roof (in Miami) and see a small orange blob rising over the horizon for about 14 seconds.
> Cue the horde of libertarians who think that it's Ma Bell's *right* to stifle innovation and how dare anyone criticise them!
Not necessarily. It's hard to think of any single entity Ayn Rand hated more completely and thoroughly than AT&T. Ayn had no love for the post office, but she DESPISED AT&T.
When I lived in Dallas in late 2000, SBC utterly and completely sucked in places like Plano where they inherited the copper of an agrarian farm town, and stapled tens of thousands of new suburbanites onto concentrators that screwed up modems and made DSL nearly impossible to get. I spent the better part of a weekend checking out apartment after apartment where DSL just wasn't available, and dialup was limited to 24k, before giving up and settling on cable internet for the summer. In contrast, BellSouth was pretty good in Miami -- I had DSL back in late '98, and knew people from Hurricane Andrew's Ground Zero who had fiber to the pod and straight-up 100mbit ethernet to their homes (actual throughput was nowhere near that, but I'm pretty sure this is the infrastructure that eventually mutated into Uverse).
> in reality having three carriers for Wireless in this country is a bad thing.
Considering that AT&T and Verizon both have long, sordid histories of selling some of the most locked-down and crippled phones ever seen on planet Earth, how exactly is the independent existence of Sprint and T-Mobile a bad thing?
Sprint itself has been fairly non-toxic, but even they refuse to allow customers to use any phone not personally approved and sold by them. T-Mobile is the sole carrier for users who value being able to buy any phone they damn well please, and get a discount in lieu of a crippled subsidized phone on top of it.
Yes, there are smaller carriers... but T-Mobile is the only one that's compatible with arbitrary imported phones, and Sprint's the only one that really, truly works nationwide and has fast data service in places besides big cities and the highways connecting them.
Horses are full of shit (which is almost pure carbon), and excrete it almost nonstop. I struggle to imagine how people used to endure living in cities like New York London, and Berlin 125 years ago when the streets were literally covered in it. Personally, if forced to choose between living in a city where the streets were ankle-deep in horse shit and living in some 1960s Soviet-era East German or Slovak industrial town with diesel buses and Ladas/Skodas/Trabants, I'd happily take the diesel fumes in a city devoid of horses.
Yeah, but if you pumped seawater across a pipeline to create ponds in Arizona, the water would quickly evaporate and leave the salt behind. Keep adding more seawater, allow more to evaporate, and eventually, it would be too salty for algae to survive.
Exactly. Red-light cameras were pushed HARD in Florida by claiming they'd reduce red-light running, but 99.7% of the actual tickets end up getting issued for rolling right turns, or coming to a complete stop with the front vertical plane of the car not 100% behind the white line painted SO FAR BACK from the intersection,drivers on wet roads end up in a position where they literally have about 200 milliseconds to figure out whether they have enough room to come to a complete stop behind the white line painted about a hundred feet back from the actual intersection without hydroplaning and spinning out of control, or trying to make it completely across the intersection to another point ridiculously far beyond the actual intersection.
The fact that there are no mandatory state-dictated standards for yellow-light timing and/or white-stripe placement makes the whole thing an even bigger farce. Actually, that's not quite right... there ARE standards for timing and geometry, but they only apply to state-road intersections... intersections that never have red-light cameras anyway because FDOT knows they're bullshit and doesn't even want to waste its time screwing with them. It's cash-starved municipalities that go crazy putting them everywhere, casting their nets as far and aggressively as they can with the tightest timings and most widely-spaced stripes the consultants leasing them the cameras and issuing the tickets tell them they can get away with. In fact, FDOT won't even allow red-light cameras on state property.
How bad is it in Florida? In many cases, the red-light cameras are now costing the municipalities money, because any ticket issued by them can be trivially challenged on technical merit and get thrown out of court with basically 100% success, often without even requiring an attorney. At least one judge (in Broward, I think) cleared his entire docket and dismissed every outstanding ticket issued by a municipality over the prior ~year because LITERALLY 100% were getting successfully thrown out, and the City's legal argument for pursuing them can be loosely summarized as "our contract with the camera company requires us to cooperate, and we really need the money." I don't remember the exact argument, but it was something along those lines, and was so egregiously bad, the judge threatened to hold the City's attorney and its elected officials in contempt of court if it kept wasting his time with tickets that couldn't stand up to even the most trivial legal challenge.
The fact is, actual honest-to-god "the light is red, and I'm going to intentionally proceed through the intersection anyway" offenses are almost *unheard* of in the US. At the end of the day, it's probably the #1 greatest cultural taboo in America. Americans will sit at a red light at 3am on a deserted 6-lane road for 5 minutes. We'll spend 10 minutes backing up and moving forward in a roomba-like forward-facing figure-8 pattern shifting across 2 or 3 lanes trying to trigger broken sensors, and do a U-turn over a curb and raised median if we believe that the light really, truly, is never going to turn green... but actually proceeding forward through a red light? Never.
It's not a matter of locked-vs-unlocked. It's a matter of frequency. Specifically, AT&T uses 850MHz and 1900MHz for UMTS/3G. Most other countries use 1900MHz and 2100MHz for UMTS/3G. T-Mobile's spectrum happens to be 1700MHz and 2100MHz. IPhone can't do 1700MHz, ergo an iPhone -- even unlocked -- on T-Mobile is never going to get better than EDGE.
It's not a permanent situation. Chipsets supporting T-Mobile's frequencies have been around for at least a few years. The only real question is whether the crop of iPhones coming out this summer will use them. Up to now, the decision to NOT use them has presumably been dictated by AT&T's desire to maximize the iPhone's incompatibility with T-Mobile, supported by the slightly lower cost of chips that don't support 1700MHz (savings of cents, not dollars) and Apple's complicity since T-Mobile/US is the only carrier on earth that uses 1700MHz.
That said, if the deal looks certain to go through, expect to see the iPhone 5.1 by Christmas, with little changed besides added support for 1700MHz. Unless there's a bizarre hole in Qualcomm's chipset family that leaves them with one variant that can do all world & American UMTS frequencies, but not CDMA2000, and another variant that can do AT&T and international UMTS frequencies and CDMA2000, but not 1700 UMTS, adding 1700MHz to an existing design is barely a parts substitution at the manufacturing level.
As far as I know, 50hz NTSC doesn't exist as a broadcast standard, but 60hz PAL most certainly DOES -- in Brazil. It's 704x480@60 fields/second, but uses PAL for color modulation instead of NTSC. There's nothing inherently 50/60-hz about PAL or NTSC.
I'm not 100% sure, but I think that back in the VHS era, Americans who wanted to watch European tapes could buy a $30 box that changed the color encoding from PAL to NTSC, but left the resolution and field rate unchanged, so you could connect a European VCR's composite output to an American TV through the box as long as the TV had sufficient vertical-hold adjustment range to tweak the field rate down to 50hz.
What's really sad is that 15 years ago, I could point my Palm at someone and trade contact info by pressing and holding a button. To this day, most Android phones STILL can't properly do bluetooth OBEX... and even if they could, I doubt whether they could exchange contact info with an iPhone.
> It is the very reason that it is and will remain a second rate gaming platform.
No, Dalvik's shitty garbage collection mechanism is. More Java developers writing Android apps get nailed by the fact that its GC stumbles over things that have been a total non-issue with desktop Java since 1.4 or 1.5 than anything else. Well, that, and the fact that as far as Eclipse is concerned, Dalvik Isn't Java (so things developers subconsciously expect to work, because they Just Work transparently for Java, don't work with Dalvik).
Any universe where Netbeans, Visual Studio, Eclipse, Microsoft Word (for anything more complicated than a single-page business letter that could be done via email), video editing, video encoding (something that can bring even the mightiest i7 to its knees; try multipass h.264 sometime...), or 3D rendering occurs. In other words, the desktop of anybody who's a content creator instead of a mere content consumer.
Back when you first started college, "video editing" involved 720x480 or 720x540 interlaced 50 or 60hz video. Now it involves 1920x1080 progressive video. Today, you could literally fill a terrabyte hard drive with an hour or two of raw, pre-encoded video. Modern DSLRs with HDR take pictures that are individually bigger (in bytes) than most CF cards that existed prior to ~7 years ago. Try editing a 12 megapixel HDR image on a PC using Photoshop on a PC with only a gig or two of ram. It's not fun.
As for people who "just want email and internet", well... Flash. Enough said.
The most high-end, hacked and rooted Android phones existing today can *barely* handle Flash in its raw, undigested, real PC form without gagging. "Works" is not the same as "runs well". Hardware-wise, a current top of the line Android phone is roughly comparable to a 500MHz Pentium 3 with 512mb, Windows 2000, and a $12 piece of shit videocard that somehow managed to have onboard MPEG-2 video acceleration anyway.
> the AGC used on the Apollo missions was considered a "real computer", and today's phones could emulate that entire system without breaking a sweat...
The difference is, the AGC made use of lots of ballistics data that was precomputed offline by mainframes and carried onboard via Hopi-woven core memory -- more megabits on a single mission, in one place at one time, than the sum total that had ever previously existed on earth. An i7 could calculate it on the fly. Had Apollo run into really, truly novel problems requiring realtime navigation that deviated significantly from the original plan (assuming it had enough fuel to allow it), the astronauts would have been fucked, because their computers wouldn't have been able to handle that use case at all.
> It's not something an enthusiast can knock out in their home workshop
Depends. Start with a FPGA eval board (about $80-200 on eBay... search ebay for nexys2), and you can do some pretty damn impressive ad-hoc faux-ASIC work in your home workshop. The only limiting factor is the fact that FPGAs are cost-prohibitive for most non-prototyping use... but if you don't *totally* care about cost and you want to achieve something that would be commercially impossible to sell (for legal, rather than technical, reasons), FPGAs are a gift from ${deity}, because Hollywood can't do a damn thing to stop you. You can even use a high-end FPGA to outright clone things like microcontrollers. Fer god's sake, people have used relatively LOW-end FPGAs to clone the Amiga 500's custom chips down to the gate and register. I give it another year or to, max, until open-source HDCP-compliant (in the sense it will work, not in the sense of "authorized") hardware designs start hitting the net now that the key is known. HDCP would have been a bitch to crack by brute force, but now that the master key is known, the bare-metal hardware needed to circumvent it isn't all that major.
2007 is around the time Nokia *finally* started to support EDGE and frequencies besides 1900/2100 on their high end phones. Circa 2004/2005, *none* of Nokia's high-end flagship phones could do EDGE. It was almost like it was against their religion or something. They were 1900/2100 UMTS and GPRS *only*.
By the way, unless something has radically changed over the past 2 or 3 years, Australia is a major CDMA market, too, as is New Zealand, India, China, Singapore, Brazil, Chile, Canada, Mexico, and a few other places as well.
> That sort of attitude is kinda what lost them the US smartphone market
No, what lost them the US smartphone market was a decision to quit making high-end CDMA phones, and an apparent inability (possibly patent-related) to make GSM phones capable of EDGE. Circa 2006, that basically meant Nokia's best phones were useless GPRS paperweights in the United States. And despite the availability of chipsets capable of both 850/1900 and 1700/2100 UMTS, Nokia took its sweet time supporting *anything* besides 1900/2100 UMTS on its high-end phones.
Distance from Boston to New York, and New York to Washington: aproximately 226 miles.
Distance from Miami to Orlando: 234 miles via most likely all-HSR route (north along I-95 to Port Canaveral, west to MCO along the Beeline Expressway)
Distance from Miami to Tampa: about 230 miles via Amtrak's current route, about 315 miles along brand new all-HSR route (continuing west from MCO to downtown Tampa)
For further comparison with Acela:
Washington Union Station to Baltimore: 38 miles Philadelphia to Wilmington, Delaware: 29 miles
Disney to Lakeland, Florida: 36 miles Tampa/Ybor City to Lakeland, Florida: 35 miles
Orlando is kind of an anomaly because it has so many high-value destinations that each merit their own stations. The biggest single problem with every proposal for HSR in Florida is the fact that they have the construction order backwards, partly due to the fixation with "true HSR". Any viable passenger rail network in Florida *must* include Orlando and Miami, and would be insane to NOT include Tampa. The problem is, the 60 miles between Boca Raton and Miami will be *ungodly* expensive to build as true HSR, because they'll literally have to be elevated 80% of the way, just like I-95 is. And somehow manage to deal with the I-95/I-595 interchange, which is so tightly interwoven and stacked, even FDOT's planners have described the task of trying to thread HSR tracks through it within the existing right of way as "nearly impossible" (it's hard to describe, but if you drive south on 95 and observe the area between Broward Boulevard and Griffin Road, or view it through Google Earth, you'll easily see why). So, supporters push for Tampa-Orlando, on the theory that it's more affordable. The problem is, "true" HSR between Orlando and Tampa only really matters if it's HSR all the way to Miami, which brings us back to square one.
> This whole tax from the state it comes from/tax from the state you live in needs to be decided (federally) and set in stone once and for all.
It was. The Supreme Court ruled that states have the authority to tax their own residents, but can't make companies from other states/countries be their unpaid tax collectors and do the job for them, nor can they compel those same companies to cooperate with their own tax-collection efforts.
Here's an example to illustrate it. Suppose that in 2000, Russia passed a law establishing a fine of 1 Euro for each use of the phrase, "In Soviet Russia" in a post to any internet forum, and made the forum's owner responsible for paying it. Ten years later, CmdrTaco opens a satellite office for Slashdot in Moscow, and starts sending checks to Russia for uses of the phrase made after that date. A few months later, he gets a tax bill demanding 400 million Euros for all the posts made after 2000, but before the office was opened and payments started to be made (and before they had a legal way to force compliance going forward). At the end of the day, Russia is no more or less entitled to fine/tax American companies with no business presence in Russia than Texas is entitled to do to a company incorporated in Washington, California, Delaware, or New York.
Bzzzt. Sorry, nice try, but wrong answer. Zipcodes don't correlate to municipal taxing boundaries. They're usually *close*, but legally that's not good enough. Just to give one example, the city of South Miami, Florida is a "patchwork city". After the first real estate crash in Florida (during the 1920s, when the combination of the Great Depression and 1926 Hurricane more or less destroyed South Florida physically AND economically), the city of South Miami went bankrupt and couldn't provide municipal services any longer. Some influential residents got the Florida legislature to allow individual landowners to secede from such municipalities. Many did. As a result, to this day, not even the city's police and fire departments can tell for sure whether or not any given square foot of land is or is not within the city limits. The net result is that nothing short of a geocoded database that resolves down to every street address in America can definitively assign every order to its proper taxing authorities. And even then, the database could become invalidated if a city changes a law and fails to notify whomever maintains the database.
It gets worse when you consider how the IRS (if it were tasked with the job) would almost certainly go about doing it -- they'd "privatize" it to a third-party that would turn around and charge anyone smaller than Wal Mart or Amazon $3 to do a "taxing-district analysis" (kind of like how when you renew your license plates in Miami, you're going to get hit for a minimum of $3 above and beyond the alleged cost of the plates/sticker themselves, because you MUST renew through a private "tag agency", and there's literally no way to pay the renewal fee that doesn't involve payment of ADDITIONAL fees for the "service" of collecting your payment and handing you the sticker to put on your plate (or the plate itself)).
> The vast majority of this country is relatively low population farmland,
The vast majority of area might be farmland or desert, but the vast majority of Americans don't actually live there. In a state like Florida, roughly 65% of the population lives within 10 miles of the Atlantic Ocean or Gulf of Mexico. Count residents who live within 10 miles of the Mouse, and the figure jumps up to about 90%. Ditto, for California. LA might Sprawl over an area more than a hundred miles wide, and almost as large from north to south, but just about everyone else lives in a 2-10 mile band around San Francisco Bay, down the Central Valley, or a small blob squished between the Pacific, mountains, and Mexico at the tail end.
To be economically viable in the medium-term, we need to forget about trying to build a single True HSR Line between the downtowns of two big cities with 3 or 4 stations in between, and instead look to the way Interstates were designed -- a few hardcore mainlines, networked with other roads that are slower, but provide seamless connectivity. By all means, spend whatever it takes between LA and San Jose, as long as the same tracks can be used by trains running on slower tracks to get to LA from San Diego, to San Jose from Sacramento, and more trains stopping at smaller towns in between that the bulk of passengers heading from LA to San Francisco don't care about and blast through at full speed without stopping. As the passenger base grows, the slower shared tracks along the periphery can be replaced by dedicated passenger tracks.
Look at some of the most important roads in America: I-80, I-95, I-75, and I-10. None of them were built from end to end in a single act. Every one of them had gaps temporarily filled in by lesser-grade roads until the past decade (technically, I-95 still has a gap in the NYNJ area, though it's more a gap in name and politics than actual road connectivity). How useful would ANY of those roads been if they had been literally separated from the rest of the nation's road network by concrete loading docks and platforms? How many families would have bought a second "Interstate Car" and kept it garaged on the other side of a concrete barrier, for exclusive use on Interstate Highways? Yet, that's exactly the ridiculous plan being pushed by many HSR supporters. They're willing to risk uselessness in the name of useless paper perfection.
HSR is good, but not when it's built at the expense of a passenger rail network that's genuinely useful, and affordable enough to build out to where people want to go. The golden range for rail travel is 200 to 400-500 miles. At that range, you can almost make up the speed difference by running additional nonstop trains between the most popular endpoints for a fraction of what it would cost to go 100% HSR. For the trains that DO stop more frequently, the time lost to dwell time and acceleration eats up most of the 150-180mph HSR advantage over 110mph ISR anyway.
> Another hurdle is the fact that we Americans love our cars.
We also love our laptops, internet connectivity, and fine dining -- all of which trains accommodate nicely, and driving usually doesn't (at least, when you're the one doing the driving). In America, the sweet spot for passenger rail is 200-~400 miles... nobody really wants to burn 4-6 hours driving, but it's not far enough to be worth the misery of flying. Have a rental car center at the station (or put the station near a major airport with direct connectivity to its ground transportation center), and the availability/suitability of local public transportation becomes a non-issue.
Couriers like UPS and FedEx might eventually start using trains for package shipments, but frankly both companies have so much invested in their own infrastructure right now, real use is more likely to come from small, regional startups in places like Florida that tag along with regional passenger rail than from big companies like UPS and FedEx. Either way, the most likely such companies would be those that provide same-day intercity package shipment, and undercut FedEx/UPS by letting customers be their own delivery system.
Ironically, transportation of stuff that's bulky and heavy (but under 50lbs, and not SO bulky that a single person can't carry it), but not particularly fragile, is one of Amtrak's best (and least-advertised) services. Heading to college and want to ship 300 pounds of books in a dozen boxes? As long as no single box exceeds 50lbs or the volume of a steamer trunk, and you have the ability to get it to and from the station at both ends, there's no cheaper way to ship them across the country, nor better timing... drop off the boxes on Monday morning in Miami, fly to $wherever on Monday afternoon, and pick up your boxes as early as Tuesday (New York or DC), Wednesday (Chicago and most of the US east of the Mississippi), Thursday (just about everywhere else besides the west coast), or Friday. And pay roughly the same amount of cash you would have paid to ship just two of the boxes via FedEx or UPS Ground (the price is based on the total weight, even if it's multiple boxes). They can even deal with heavier stuff, as long as you're personally able to get it loaded onto a pallet without assistance from them. I wouldn't use Amtrak to ship anything breakable, but for stuff like books and boxed clothing, they're pretty much unbeatable.
> I wonder if we have the wherewithal to keep paying for it long enough for people to really start using it
That's why the US needs to forget about 100% virgin flawless HSR for now, and start with improved regular tracks that are occasionally (but not heavily) shared with freight and trains that run 80-110mph. Getting to 110mph is fairly cheap -- roughly $1-5 million per mile (not counting bridges) through areas where there's an existing rail corridor with existing trackbed that's still in decent condition and just needs to have new tracks laid on top of it & some long-overdue maintenance work on the rest. By sticking with cheap and frequent service that's dispatched to favor time-sensitive passenger trains, you buy time for consumer demand to grow before you start to take on the *really* expensive construction projects. The big thing is to ensure that the service is useful enough to start attracting real customers on Day One, without being burdened by a crushing debt load that makes further expansion politically impossible.
When the day comes that rebuilding as True HSR is really necessary, people won't *care* how much it's going to cost, because it'll be something they and everyone they know uses and depends upon. Aside from environmentalists and NIMBYs looking for an excuse, how many people *really* complain about the cost of major freeway-widening projects? Almost none, because everyone just wants it done, and the want it done *now*. The key is keeping the cost of passenger rail construction low enough to allow demand to grow to that point in the first place.
Slashdot Mirror
^^^ argh. This is what happens when you post replies to Slashdot while eating breakfast, before the caffeine kicks in.
I went back and looked at my code from a few months ago. I don't have time to re-study it now to figure out what went wrong (I'm already late for work), but from what I remember, I either couldn't get it to do Cipher.getInstance() without throwing an exception, I had problems because Android only includes a partial implementation of Bouncycastle, and trying to drop the entire jarfile caused classloading problems, or maybe it just didn't work. I just remember giving up at some point and deciding it was a lost cause.
Here's my source, if anyone wants to play with it:
Cipher rsa = null;
byte[] password = (new String("passwordtest")).getBytes();
try { //...
rsa = Cipher.getInstance("RSA/ECB/NOPADDING");
rsa.init(Cipher.ENCRYPT_MODE, RsaKey.getInstance());
byte[] fakeHash = rsa.doFinal(password);
for (int x=0; x < password.length; x++) {
log.i(x + " = " + Integer.toString(password[x] & 0xff, 16));
}
for (int x=0; x < fakeHash.length; x++) {
log.i(x + " = " + Integer.toString(fakeHash[x] & 0xff, 16));
}
}
This is a more or less verbatim repeat of what I said to someone else, but it merits repeating here because your post addresses my response's topic head-on: the fundamental problem with that approach is that RSA and AES both include an element of random variation in the encrypted text. In other words, given a plaintext password and a public key, you can't assume that the encrypted output you get THIS time will be identical to the encrypted output you'll get NEXT time (or from a different implementation of the encryption algorithm). The encrypted output will decrypt with the private key just fine, but basically this means you can't treat the encrypted output like a reversible hash that just happens to take a lot longer to compute.
From what I remember from my research into this specific issue a few months ago, it's technically possible to write an implementation of RSA's encryption algorithm that introduces exactly zero bytes of random salt into the encrypted output, but I wasn't able to make it work using Bouncycastle, and got the impression that you can't do it with Microsoft's implementations or any other commonly-used implementation for Java, either.
AFAIK, there's no way at all to encrypt using AES without random entropy. It's an inherent part of the algorithm. In RSA's case, it's technically possible because it was tacked on as an afterthought, and things encrypted using the original algorithm end up looking like something encrypted using the new algorithm with zero bytes of random salt.
I'm not sure about other asymmetric algorithms. I remember hearing a lot about Elgamal back when I was in college, but ever since RSA's patent issues went away and AES became officially blessed by the US federal government, pretty much everything besides RSA and AES seem to have fallen by the wayside.
I had a similar idea a few months ago. I thought about using asymmetric encryption, and treating the values encrypted with the public key like hashes -- storing the encrypted values in the database, and keeping the private key safe and used only for offline decryption if necessary to fix a bug or something. Then I discovered the big problem -- AES and RSA include an element of randomness in their output. In other words, given the same input value and public key, you will NOT necessarily get the same encrypted output every time. Whatever you get WILL decrypt properly with the private key, but you can't assume that the encrypted output of a given input + public key you got this time will be identical to the encrypted output you'll get next time.
From what I remember, eliminating that randomness is impossible with AES (it's an inherent part of the algorithm). It might be possible with RSA *if* you had your own implementation of the encryption algorithm, but all the common implementations used by normal people (Microsoft, Bouncycastle, IBM, etc) do it. I'm not sure about Elgamal, or other algorithms that have fallen out of favor since RSA's patent expired and AES became the official standard of the US Federal Government.
I'm going to the launch, too. Realistically, how close do you have to BE in order to really experience it (hear the engines, be able to see it without binoculars, or at least without expensive binoculars, etc)? The closest I've ever come to seeing a shuttle launch in person was back around '94, when there was a night launch on a rare cloudless night, and I was able to stand on the roof (in Miami) and see a small orange blob rising over the horizon for about 14 seconds.
Ah, if he's saying that "three is worse than four", I agree. I read his post as "three are too many, and we'd be better off with two."
> Cue the horde of libertarians who think that it's Ma Bell's *right* to stifle innovation and how dare anyone criticise them!
Not necessarily. It's hard to think of any single entity Ayn Rand hated more completely and thoroughly than AT&T. Ayn had no love for the post office, but she DESPISED AT&T.
When I lived in Dallas in late 2000, SBC utterly and completely sucked in places like Plano where they inherited the copper of an agrarian farm town, and stapled tens of thousands of new suburbanites onto concentrators that screwed up modems and made DSL nearly impossible to get. I spent the better part of a weekend checking out apartment after apartment where DSL just wasn't available, and dialup was limited to 24k, before giving up and settling on cable internet for the summer. In contrast, BellSouth was pretty good in Miami -- I had DSL back in late '98, and knew people from Hurricane Andrew's Ground Zero who had fiber to the pod and straight-up 100mbit ethernet to their homes (actual throughput was nowhere near that, but I'm pretty sure this is the infrastructure that eventually mutated into Uverse).
> in reality having three carriers for Wireless in this country is a bad thing.
Considering that AT&T and Verizon both have long, sordid histories of selling some of the most locked-down and crippled phones ever seen on planet Earth, how exactly is the independent existence of Sprint and T-Mobile a bad thing?
Sprint itself has been fairly non-toxic, but even they refuse to allow customers to use any phone not personally approved and sold by them. T-Mobile is the sole carrier for users who value being able to buy any phone they damn well please, and get a discount in lieu of a crippled subsidized phone on top of it.
Yes, there are smaller carriers... but T-Mobile is the only one that's compatible with arbitrary imported phones, and Sprint's the only one that really, truly works nationwide and has fast data service in places besides big cities and the highways connecting them.
Horses are full of shit (which is almost pure carbon), and excrete it almost nonstop. I struggle to imagine how people used to endure living in cities like New York London, and Berlin 125 years ago when the streets were literally covered in it. Personally, if forced to choose between living in a city where the streets were ankle-deep in horse shit and living in some 1960s Soviet-era East German or Slovak industrial town with diesel buses and Ladas/Skodas/Trabants, I'd happily take the diesel fumes in a city devoid of horses.
Yeah, but if you pumped seawater across a pipeline to create ponds in Arizona, the water would quickly evaporate and leave the salt behind. Keep adding more seawater, allow more to evaporate, and eventually, it would be too salty for algae to survive.
Exactly. Red-light cameras were pushed HARD in Florida by claiming they'd reduce red-light running, but 99.7% of the actual tickets end up getting issued for rolling right turns, or coming to a complete stop with the front vertical plane of the car not 100% behind the white line painted SO FAR BACK from the intersection,drivers on wet roads end up in a position where they literally have about 200 milliseconds to figure out whether they have enough room to come to a complete stop behind the white line painted about a hundred feet back from the actual intersection without hydroplaning and spinning out of control, or trying to make it completely across the intersection to another point ridiculously far beyond the actual intersection.
The fact that there are no mandatory state-dictated standards for yellow-light timing and/or white-stripe placement makes the whole thing an even bigger farce. Actually, that's not quite right... there ARE standards for timing and geometry, but they only apply to state-road intersections... intersections that never have red-light cameras anyway because FDOT knows they're bullshit and doesn't even want to waste its time screwing with them. It's cash-starved municipalities that go crazy putting them everywhere, casting their nets as far and aggressively as they can with the tightest timings and most widely-spaced stripes the consultants leasing them the cameras and issuing the tickets tell them they can get away with. In fact, FDOT won't even allow red-light cameras on state property.
How bad is it in Florida? In many cases, the red-light cameras are now costing the municipalities money, because any ticket issued by them can be trivially challenged on technical merit and get thrown out of court with basically 100% success, often without even requiring an attorney. At least one judge (in Broward, I think) cleared his entire docket and dismissed every outstanding ticket issued by a municipality over the prior ~year because LITERALLY 100% were getting successfully thrown out, and the City's legal argument for pursuing them can be loosely summarized as "our contract with the camera company requires us to cooperate, and we really need the money." I don't remember the exact argument, but it was something along those lines, and was so egregiously bad, the judge threatened to hold the City's attorney and its elected officials in contempt of court if it kept wasting his time with tickets that couldn't stand up to even the most trivial legal challenge.
The fact is, actual honest-to-god "the light is red, and I'm going to intentionally proceed through the intersection anyway" offenses are almost *unheard* of in the US. At the end of the day, it's probably the #1 greatest cultural taboo in America. Americans will sit at a red light at 3am on a deserted 6-lane road for 5 minutes. We'll spend 10 minutes backing up and moving forward in a roomba-like forward-facing figure-8 pattern shifting across 2 or 3 lanes trying to trigger broken sensors, and do a U-turn over a curb and raised median if we believe that the light really, truly, is never going to turn green... but actually proceeding forward through a red light? Never.
It's not a matter of locked-vs-unlocked. It's a matter of frequency. Specifically, AT&T uses 850MHz and 1900MHz for UMTS/3G. Most other countries use 1900MHz and 2100MHz for UMTS/3G. T-Mobile's spectrum happens to be 1700MHz and 2100MHz. IPhone can't do 1700MHz, ergo an iPhone -- even unlocked -- on T-Mobile is never going to get better than EDGE.
It's not a permanent situation. Chipsets supporting T-Mobile's frequencies have been around for at least a few years. The only real question is whether the crop of iPhones coming out this summer will use them. Up to now, the decision to NOT use them has presumably been dictated by AT&T's desire to maximize the iPhone's incompatibility with T-Mobile, supported by the slightly lower cost of chips that don't support 1700MHz (savings of cents, not dollars) and Apple's complicity since T-Mobile/US is the only carrier on earth that uses 1700MHz.
That said, if the deal looks certain to go through, expect to see the iPhone 5.1 by Christmas, with little changed besides added support for 1700MHz. Unless there's a bizarre hole in Qualcomm's chipset family that leaves them with one variant that can do all world & American UMTS frequencies, but not CDMA2000, and another variant that can do AT&T and international UMTS frequencies and CDMA2000, but not 1700 UMTS, adding 1700MHz to an existing design is barely a parts substitution at the manufacturing level.
>I'll bet that East Japan was PAL back then.
As far as I know, 50hz NTSC doesn't exist as a broadcast standard, but 60hz PAL most certainly DOES -- in Brazil. It's 704x480@60 fields/second, but uses PAL for color modulation instead of NTSC. There's nothing inherently 50/60-hz about PAL or NTSC.
I'm not 100% sure, but I think that back in the VHS era, Americans who wanted to watch European tapes could buy a $30 box that changed the color encoding from PAL to NTSC, but left the resolution and field rate unchanged, so you could connect a European VCR's composite output to an American TV through the box as long as the TV had sufficient vertical-hold adjustment range to tweak the field rate down to 50hz.
What's really sad is that 15 years ago, I could point my Palm at someone and trade contact info by pressing and holding a button. To this day, most Android phones STILL can't properly do bluetooth OBEX... and even if they could, I doubt whether they could exchange contact info with an iPhone.
> It is the very reason that it is and will remain a second rate gaming platform.
No, Dalvik's shitty garbage collection mechanism is. More Java developers writing Android apps get nailed by the fact that its GC stumbles over things that have been a total non-issue with desktop Java since 1.4 or 1.5 than anything else. Well, that, and the fact that as far as Eclipse is concerned, Dalvik Isn't Java (so things developers subconsciously expect to work, because they Just Work transparently for Java, don't work with Dalvik).
> In what universe is that not enough?
Any universe where Netbeans, Visual Studio, Eclipse, Microsoft Word (for anything more complicated than a single-page business letter that could be done via email), video editing, video encoding (something that can bring even the mightiest i7 to its knees; try multipass h.264 sometime...), or 3D rendering occurs. In other words, the desktop of anybody who's a content creator instead of a mere content consumer.
Back when you first started college, "video editing" involved 720x480 or 720x540 interlaced 50 or 60hz video. Now it involves 1920x1080 progressive video. Today, you could literally fill a terrabyte hard drive with an hour or two of raw, pre-encoded video. Modern DSLRs with HDR take pictures that are individually bigger (in bytes) than most CF cards that existed prior to ~7 years ago. Try editing a 12 megapixel HDR image on a PC using Photoshop on a PC with only a gig or two of ram. It's not fun.
As for people who "just want email and internet", well... Flash. Enough said.
The most high-end, hacked and rooted Android phones existing today can *barely* handle Flash in its raw, undigested, real PC form without gagging. "Works" is not the same as "runs well". Hardware-wise, a current top of the line Android phone is roughly comparable to a 500MHz Pentium 3 with 512mb, Windows 2000, and a $12 piece of shit videocard that somehow managed to have onboard MPEG-2 video acceleration anyway.
> the AGC used on the Apollo missions was considered a "real computer", and today's phones could emulate that entire system without breaking a sweat...
The difference is, the AGC made use of lots of ballistics data that was precomputed offline by mainframes and carried onboard via Hopi-woven core memory -- more megabits on a single mission, in one place at one time, than the sum total that had ever previously existed on earth. An i7 could calculate it on the fly. Had Apollo run into really, truly novel problems requiring realtime navigation that deviated significantly from the original plan (assuming it had enough fuel to allow it), the astronauts would have been fucked, because their computers wouldn't have been able to handle that use case at all.
> It's not something an enthusiast can knock out in their home workshop
Depends. Start with a FPGA eval board (about $80-200 on eBay... search ebay for nexys2), and you can do some pretty damn impressive ad-hoc faux-ASIC work in your home workshop. The only limiting factor is the fact that FPGAs are cost-prohibitive for most non-prototyping use... but if you don't *totally* care about cost and you want to achieve something that would be commercially impossible to sell (for legal, rather than technical, reasons), FPGAs are a gift from ${deity}, because Hollywood can't do a damn thing to stop you. You can even use a high-end FPGA to outright clone things like microcontrollers. Fer god's sake, people have used relatively LOW-end FPGAs to clone the Amiga 500's custom chips down to the gate and register. I give it another year or to, max, until open-source HDCP-compliant (in the sense it will work, not in the sense of "authorized") hardware designs start hitting the net now that the key is known. HDCP would have been a bitch to crack by brute force, but now that the master key is known, the bare-metal hardware needed to circumvent it isn't all that major.
(*blush*) Whoops. I didn't know that.
2007 is around the time Nokia *finally* started to support EDGE and frequencies besides 1900/2100 on their high end phones. Circa 2004/2005, *none* of Nokia's high-end flagship phones could do EDGE. It was almost like it was against their religion or something. They were 1900/2100 UMTS and GPRS *only*.
By the way, unless something has radically changed over the past 2 or 3 years, Australia is a major CDMA market, too, as is New Zealand, India, China, Singapore, Brazil, Chile, Canada, Mexico, and a few other places as well.
> That sort of attitude is kinda what lost them the US smartphone market
No, what lost them the US smartphone market was a decision to quit making high-end CDMA phones, and an apparent inability (possibly patent-related) to make GSM phones capable of EDGE. Circa 2006, that basically meant Nokia's best phones were useless GPRS paperweights in the United States. And despite the availability of chipsets capable of both 850/1900 and 1700/2100 UMTS, Nokia took its sweet time supporting *anything* besides 1900/2100 UMTS on its high-end phones.
Distance from Boston to New York, and New York to Washington: aproximately 226 miles.
Distance from Miami to Orlando: 234 miles via most likely all-HSR route (north along I-95 to Port Canaveral, west to MCO along the Beeline Expressway)
Distance from Miami to Tampa: about 230 miles via Amtrak's current route, about 315 miles along brand new all-HSR route (continuing west from MCO to downtown Tampa)
For further comparison with Acela:
Washington Union Station to Baltimore: 38 miles
Philadelphia to Wilmington, Delaware: 29 miles
Disney to Lakeland, Florida: 36 miles
Tampa/Ybor City to Lakeland, Florida: 35 miles
Orlando is kind of an anomaly because it has so many high-value destinations that each merit their own stations. The biggest single problem with every proposal for HSR in Florida is the fact that they have the construction order backwards, partly due to the fixation with "true HSR". Any viable passenger rail network in Florida *must* include Orlando and Miami, and would be insane to NOT include Tampa. The problem is, the 60 miles between Boca Raton and Miami will be *ungodly* expensive to build as true HSR, because they'll literally have to be elevated 80% of the way, just like I-95 is. And somehow manage to deal with the I-95/I-595 interchange, which is so tightly interwoven and stacked, even FDOT's planners have described the task of trying to thread HSR tracks through it within the existing right of way as "nearly impossible" (it's hard to describe, but if you drive south on 95 and observe the area between Broward Boulevard and Griffin Road, or view it through Google Earth, you'll easily see why). So, supporters push for Tampa-Orlando, on the theory that it's more affordable. The problem is, "true" HSR between Orlando and Tampa only really matters if it's HSR all the way to Miami, which brings us back to square one.
> This whole tax from the state it comes from/tax from the state you live in needs to be decided (federally) and set in stone once and for all.
It was. The Supreme Court ruled that states have the authority to tax their own residents, but can't make companies from other states/countries be their unpaid tax collectors and do the job for them, nor can they compel those same companies to cooperate with their own tax-collection efforts.
Here's an example to illustrate it. Suppose that in 2000, Russia passed a law establishing a fine of 1 Euro for each use of the phrase, "In Soviet Russia" in a post to any internet forum, and made the forum's owner responsible for paying it. Ten years later, CmdrTaco opens a satellite office for Slashdot in Moscow, and starts sending checks to Russia for uses of the phrase made after that date. A few months later, he gets a tax bill demanding 400 million Euros for all the posts made after 2000, but before the office was opened and payments started to be made (and before they had a legal way to force compliance going forward). At the end of the day, Russia is no more or less entitled to fine/tax American companies with no business presence in Russia than Texas is entitled to do to a company incorporated in Washington, California, Delaware, or New York.
> State, zip code and tax rate?
Bzzzt. Sorry, nice try, but wrong answer. Zipcodes don't correlate to municipal taxing boundaries. They're usually *close*, but legally that's not good enough. Just to give one example, the city of South Miami, Florida is a "patchwork city". After the first real estate crash in Florida (during the 1920s, when the combination of the Great Depression and 1926 Hurricane more or less destroyed South Florida physically AND economically), the city of South Miami went bankrupt and couldn't provide municipal services any longer. Some influential residents got the Florida legislature to allow individual landowners to secede from such municipalities. Many did. As a result, to this day, not even the city's police and fire departments can tell for sure whether or not any given square foot of land is or is not within the city limits. The net result is that nothing short of a geocoded database that resolves down to every street address in America can definitively assign every order to its proper taxing authorities. And even then, the database could become invalidated if a city changes a law and fails to notify whomever maintains the database.
It gets worse when you consider how the IRS (if it were tasked with the job) would almost certainly go about doing it -- they'd "privatize" it to a third-party that would turn around and charge anyone smaller than Wal Mart or Amazon $3 to do a "taxing-district analysis" (kind of like how when you renew your license plates in Miami, you're going to get hit for a minimum of $3 above and beyond the alleged cost of the plates/sticker themselves, because you MUST renew through a private "tag agency", and there's literally no way to pay the renewal fee that doesn't involve payment of ADDITIONAL fees for the "service" of collecting your payment and handing you the sticker to put on your plate (or the plate itself)).
> The vast majority of this country is relatively low population farmland,
The vast majority of area might be farmland or desert, but the vast majority of Americans don't actually live there. In a state like Florida, roughly 65% of the population lives within 10 miles of the Atlantic Ocean or Gulf of Mexico. Count residents who live within 10 miles of the Mouse, and the figure jumps up to about 90%. Ditto, for California. LA might Sprawl over an area more than a hundred miles wide, and almost as large from north to south, but just about everyone else lives in a 2-10 mile band around San Francisco Bay, down the Central Valley, or a small blob squished between the Pacific, mountains, and Mexico at the tail end.
To be economically viable in the medium-term, we need to forget about trying to build a single True HSR Line between the downtowns of two big cities with 3 or 4 stations in between, and instead look to the way Interstates were designed -- a few hardcore mainlines, networked with other roads that are slower, but provide seamless connectivity. By all means, spend whatever it takes between LA and San Jose, as long as the same tracks can be used by trains running on slower tracks to get to LA from San Diego, to San Jose from Sacramento, and more trains stopping at smaller towns in between that the bulk of passengers heading from LA to San Francisco don't care about and blast through at full speed without stopping. As the passenger base grows, the slower shared tracks along the periphery can be replaced by dedicated passenger tracks.
Look at some of the most important roads in America: I-80, I-95, I-75, and I-10. None of them were built from end to end in a single act. Every one of them had gaps temporarily filled in by lesser-grade roads until the past decade (technically, I-95 still has a gap in the NYNJ area, though it's more a gap in name and politics than actual road connectivity). How useful would ANY of those roads been if they had been literally separated from the rest of the nation's road network by concrete loading docks and platforms? How many families would have bought a second "Interstate Car" and kept it garaged on the other side of a concrete barrier, for exclusive use on Interstate Highways? Yet, that's exactly the ridiculous plan being pushed by many HSR supporters. They're willing to risk uselessness in the name of useless paper perfection.
HSR is good, but not when it's built at the expense of a passenger rail network that's genuinely useful, and affordable enough to build out to where people want to go. The golden range for rail travel is 200 to 400-500 miles. At that range, you can almost make up the speed difference by running additional nonstop trains between the most popular endpoints for a fraction of what it would cost to go 100% HSR. For the trains that DO stop more frequently, the time lost to dwell time and acceleration eats up most of the 150-180mph HSR advantage over 110mph ISR anyway.
> Another hurdle is the fact that we Americans love our cars.
We also love our laptops, internet connectivity, and fine dining -- all of which trains accommodate nicely, and driving usually doesn't (at least, when you're the one doing the driving). In America, the sweet spot for passenger rail is 200-~400 miles... nobody really wants to burn 4-6 hours driving, but it's not far enough to be worth the misery of flying. Have a rental car center at the station (or put the station near a major airport with direct connectivity to its ground transportation center), and the availability/suitability of local public transportation becomes a non-issue.
Couriers like UPS and FedEx might eventually start using trains for package shipments, but frankly both companies have so much invested in their own infrastructure right now, real use is more likely to come from small, regional startups in places like Florida that tag along with regional passenger rail than from big companies like UPS and FedEx. Either way, the most likely such companies would be those that provide same-day intercity package shipment, and undercut FedEx/UPS by letting customers be their own delivery system.
Ironically, transportation of stuff that's bulky and heavy (but under 50lbs, and not SO bulky that a single person can't carry it), but not particularly fragile, is one of Amtrak's best (and least-advertised) services. Heading to college and want to ship 300 pounds of books in a dozen boxes? As long as no single box exceeds 50lbs or the volume of a steamer trunk, and you have the ability to get it to and from the station at both ends, there's no cheaper way to ship them across the country, nor better timing... drop off the boxes on Monday morning in Miami, fly to $wherever on Monday afternoon, and pick up your boxes as early as Tuesday (New York or DC), Wednesday (Chicago and most of the US east of the Mississippi), Thursday (just about everywhere else besides the west coast), or Friday. And pay roughly the same amount of cash you would have paid to ship just two of the boxes via FedEx or UPS Ground (the price is based on the total weight, even if it's multiple boxes). They can even deal with heavier stuff, as long as you're personally able to get it loaded onto a pallet without assistance from them. I wouldn't use Amtrak to ship anything breakable, but for stuff like books and boxed clothing, they're pretty much unbeatable.
> I wonder if we have the wherewithal to keep paying for it long enough for people to really start using it
That's why the US needs to forget about 100% virgin flawless HSR for now, and start with improved regular tracks that are occasionally (but not heavily) shared with freight and trains that run 80-110mph. Getting to 110mph is fairly cheap -- roughly $1-5 million per mile (not counting bridges) through areas where there's an existing rail corridor with existing trackbed that's still in decent condition and just needs to have new tracks laid on top of it & some long-overdue maintenance work on the rest. By sticking with cheap and frequent service that's dispatched to favor time-sensitive passenger trains, you buy time for consumer demand to grow before you start to take on the *really* expensive construction projects. The big thing is to ensure that the service is useful enough to start attracting real customers on Day One, without being burdened by a crushing debt load that makes further expansion politically impossible.
When the day comes that rebuilding as True HSR is really necessary, people won't *care* how much it's going to cost, because it'll be something they and everyone they know uses and depends upon. Aside from environmentalists and NIMBYs looking for an excuse, how many people *really* complain about the cost of major freeway-widening projects? Almost none, because everyone just wants it done, and the want it done *now*. The key is keeping the cost of passenger rail construction low enough to allow demand to grow to that point in the first place.
For anyone wondering, the