^^^ That reminds me of another caveat -- inrush current. Powerful fans, in particular, are hard to use with generators -- even big ones. I have a Vornado fan (circa 1995) that can almost blow the bark off of a tree when it's running at full speed. My first generator (a 4-stroke 2000-watt baby generator like the one I described above) couldn't run it. I plugged it into the extension cord, turned it on (after starting the generator and letting it stabilize), and the generator literally rocked about 3 inches in the air on one side and choked to a halt as though an invisible hand just grabbed the spinning rotor and forced it to stop. The same generator was able to start a cheap window box fan... but ONLY if I quickly turned the knob from "off" to "medium" and allowed it to stabilize before turning it up to 'high'. If I went directly from "off" to "high", it would stall the generator.
The microwave oven was another thing that the generator didn't like *at all*. I tried using the microwave with generator #2 (5600-watt Craftsman). It worked, but both the microwave and generator made really bad-sounding noises (hard to describe, kind of a buzzing hum that was REALLY loud), and I decided to just forget about trying to use the microwave on generator power due to worries that it would damage the oven, the generator, or both.
That reminds me... if you're in the hurricane's path, do all your laundry now. You can run a washing machine from a generator, but even a whole-house 24-kW Generac is going to struggle with an electric dryer. I don't know about the northeast, but in Florida, clothes lines just don't work during the summer. You can leave clothes hanging on them all day, and they'll STILL be damp when the sun goes down. Post-Wilma, my coworkers and I had to bring damp clothes to the office and hang them on makeshift clotheslines between cubes to get them to dry out in the air conditioning.
After Hurricane Wilma, I had no power for almost four weeks.
Four. Fucking. Weeks.
I didn't live out in BFE, either... I lived in Coral Gables, which is about as hardcore "Central Dade County" as you can get.
That said, here's a big, huge tip for anybody who wants to be able to run a window air conditioner from a generator -- all things equal, the magic minimum is around 3,600 watts. I'd recommend 4200-4800 minimum. Why?
1. The generator's wattage is a polite fiction. The number printed on the box is roughly what it can output for about 5 minutes before Bad Things Happen. The REAL power output it can SUSTAIN is about 80% of that amount, maybe less.
2. Most generators have split-phase power, which is a nice way of saying that the big number printed by the box is kind of divided between two outlets. So your "4800-watt" generator is really more like 2400 watts (max) per outlet (which translates into about 1800 watts per outlet sustained). A small window air conditioner draws about 1200-1500 running watts, and needs about 1800 watts to start up.
Now, the half & half rule isn't quite set in stone... you can usually get away with drawing about 2400 watts (sustained) from a single outlet on a "4000-watt" generator with no load on the other outlet, but then you run into the next problem:
3. Generator run times are usually quoted at "50% load". If you have an air conditioner connected to one outlet of a 4KW generator, it's not really a "50% load", even if it's the only thing you're running. Why? Unbalanced loads make your fuel economy go WAY down. It won't quite suck down as much gas as a 100% load, but from my own experience, it'll act kind of like an 80-90% load fuel-wise. So if you're going to run a window air conditioner from a 4KW generator, you might as well plug the refrigerator (or another small air conditioner) into the other outlet and enjoy it, because at that point it will barely make a dent in your fuel use.
That said, don't go hog wild and buy a 10KW generator without a good reason. Especially not a cheap one. Most cheap generators do a really bad job of throttling down to accommodate reduced loads, and will burn almost as much gas with a nightlight as they will with a 50% load. It's a balancing act, and it's an important one, because if you're going to be feeding a generator for a few days, let alone a few weeks, a $40-50/day gas habit quickly becomes painful.
Oh, I almost forgot... there's one last catch...
4. Generators and UPSes don't get along. At all. 99.9% of the UPSes you can buy at a retail store will ignore electricity from a generator, will run 100% from the battery until it's drained, and shut down. There ARE expensive inverter-type generators that can charge a UPS, and UPSes that can charge from a cheap generator, but both are likely to cost more than it's worth spending.
4b. Generators and some DC power supplies don't get along very well, either. It's hit-or-miss, and hard to tell which power supplies are generator-unfriendly without testing them. Some will operate very, very inefficiently, and some won't work at all. The problem is that cheap (non inverter-type) generators don't output sine waves, and their "dirty" output doesn't play nicely with switching-type power supplies. You MIGHT be able to get around this by "double conversion". After Wilma, I had to power my DSL modem by plugging a 12v adapter into an outlet (which gave me a fake cigarette lighter rated at 1000mA), then plugged an inverter into it (giving me a 110v outlet), then plugged the DSL modem's power supply into the inverter. Ugly in countless ways, but it got me back online.
4c. As a corollary to 4b, most cheap generators suck at battery-charging.
The moral: if you don't need air conditioning, and can afford it, buy an inverter-type generator. They'll play nicely with power supplies (but your UPS might still get bitchy), and low-RPM expensive inverter-type generators also tend to be the quietest and most fuel efficient. Apparently, Honda makes some o
I meant, beyond the strait after it made landfall on the US side. Obviously the part under the strait itself is meant to be a tunnel. The specific obstacle I was referring to was not open water, but glacier fields where conventional civil engineering kind of falls flat on its face.
I'd be more concerned about permafrost, and possibly glaciers. Especially glaciers -- they're the one surface terrain where we're still pretty much in the stone age insofar as the construction of permanent precision infrastructure is concerned. You can't sink pilings down to the bedrock, because glaciers flow (slowly) and would eventually shear them away at "ground" level. You can't treat ice like bedrock, because pressure liquefies it and causes whatever you built on top to slowly sink. To cross a glacier field, you'd literally have to blast a canyon into the ice down to the permafrost, THEN build it on refrigerated concrete pilings to keep the ground around it frozen forever. Or maybe build a long multi-span suspension bridge whose support columns rested on concrete foundations big enough to flow the ice around them (with occasional seismic-like activity when the ice briefly prevails and manages to shove the concrete artificial island a few inches). Even if the environmentalists didn't go into convulsions, I'm sure it would be so cost-insane it would almost make sense to build the whole thing in a deep bored tunnel instead.
That said, if they can solve the civil engineering problems in a way that won't bankrupt the venture, it would be pretty cool. If nothing else, it would mean there would eventually be cheaper transportation links between Alaska and the lower 48 + Canada as well, so the state of Alaska would no longer have to budget enough food stamps for indigenous people who insist on living 500 miles from the nearest Wal-Mart and pay something like $12 for a box of Rice Krispies.
> A wave of locusts in Kansas would bump worldwide food costs up a couple of cents.
Assuming there's abundance elsewhere in the system to absorb the shock. The problem is, if you eliminate the subsidies and EVERYONE starts to farm for maximum shareholder value and return on investment, most of that worldwide capacity would eventually disappear and be wrung out of the system the same way the US mortgage industry was leveraged to the breaking point with toxic assets that COULDN'T be unloaded quickly enough to preserve necessary liquidity to avoid a collapse. It wouldn't happen in a year or two, but eventually, you WOULD end up in situations where a really bad crop failure in one or two parts of the US would be painful everywhere, because the other places that would have otherwise grown unprofitable (without subsidies) crops would be growing more profitable things instead (or be using the land to build suburban estate homes for people with horses, for example).
Growing lettuce in the Arizona desert isn't economically efficient, but it DOES help to guarantee that a major crop failure, um, wherever it is that lettuce normally gets grown will be reflected in a minor temporary price bump instead of lettuce simply becoming unavailable for weeks or months. Americans like knowing that we can go to the 'produce' section of the grocery store confident that we'll find exactly the same pristine, flawless fruits and vegetables we found last week, for more or less the same price, regardless of month, weather, or news reports. In other parts of the world, where things ARE left more up to pure market forces, that's NOT necessarily the case.
That's not to say the system is anywhere near perfect or entirely for the purpose of avoiding famine (tobacco being the obvious example). It does, however, go a long way towards making the availability of flawless fresh food something that "just works" in the background and most Americans never have to be bothered with.
China has an official government standard for micro-USB charging that basically says, "if the data lines are shorted together, the client can legitimately assume that the host is a charger capable of supplying 1.7 Amperes". By virtue of everything coming from China, that basically means that there's now a worldwide standard for the same. However, most cheap USB chargers DON'T short the pins together, because then they'd have to have a regulator with proper heatsink capable of supplying up to 1.7A continuously without melting or overheating. I believe most of the really, really cheap chargers use parts that will quickly die if you try to draw more than ~500mA continuously.
This is why, for example, a Samsung or HTC Android phone charges at least twice as fast when connected to their official chargers (and some thirdparty chargers) than it does when connected to a powered USB hub, and several orders of magnitude faster than they'll charge from a laptop's USB port. The bundled power supply shorts D+ and D-, so the phone knows it can draw 700mA (Samsung) or some comparable amount (HTC). Likewise, if the phone senses USB, but can't discern its power capabilities, it will assume the worst and limit its draw to 100mA (this is why quite a few Android phones will tread water instead of charge if you're tethered via USB... they're drawing power from the host PC, but less total current than they're using at that moment to maintain an active data connection.
Just wait until someone finally comes up with a viable standard for some hybrid between cellular and wi-fi (802.69?) that basically puts gigabit wiring back into the walls so it can feed hundreds of access points with service radii of ~10-25 feet apiece, and seamless hand-offs as users move around. Wi-fi is easy if your goal is to enable somebody with a laptop to get online at low speeds and not move around a lot. Enabling somebody to walk across the office and enjoy seamless wireless 100+mbit/sec connectivity every inch of the way is another matter entirely, and a problem companies like Cisco have barely even *started* to work on.
If you're going to go with football analogies, Apple is scoring a touchdown with two-point bonus every hour. Google, meanwhile, is massively spoiling their plans and running up the score by intercepting the ball every time Apple tries to do something and immediately kicking successful field goals from the 90-yard line. BCS is being grumpy and giving Apple lots of extra votes because it's scoring "legitimate" points while Google is gaming the system and scoring lots of points in ways BCS doesn't necessarily agree with, but at the end of the day, Google's going to go home victorious with a score of 90-24, even if Apple ends up getting invited to the best bowl game. Team Google doesn't care, because they won, stomped all over Apple, and they're getting laid tonight anyway after the victory party.
What you need is a USB CondomCable with the D+ and D- pins shorted together. No data can flow, and if the bad guys didn't bother to try and implement proper power protocol, you'll get the added satisfaction of frying THEIR hardware when your phone cranks up the juice and tries to suck down 1.7A instead of politely sipping 100mA. Just don't ever use such a cable by mistake to connect your phone to a pc or laptop belonging to yourself or a friend.
Frighteningly enough, I can see this happening. It would almost be better to plot the sail's trajectory to screw with its trajectory if it passes THROUGH the keyhole, and glide by harmlessly if Apophis MISSES the keyhole. By all means get the hardware into position to avert disaster... but if it looks like disaster isn't likely to happen, for the love of ${deity} don't go screwing with it and risk making things worse just for the sake of Doing Something.
It's kind of like theoretical weather-control experiments. If we someday know Tropical Storm Tyreesha is predicted to hit Miami on September 18th and disrupt the season finale of a popular mass-public reality TV show, do we REALLY want to risk kicking it back into the Straits of Florida and nudging it south so it passes harmlessly over the southern Everglades... then unexpectedly deviates from the models, grows into a monster category 5 hurricane over the Gulf, and destroys Houston the way Andrew destroyed Miami, instead of dumping a foot of rain on South Florida, fizzling out, sputtering back to life, and drifting north until it trips over southern Alabama as a minor category 1 hurricane?
Thought experiment: someday, a small asteroid is discovered that, if we do nothing, will strike a relatively desolate part of the earth and do about as much real damage as Tunguska. We have the ability to nudge it a bit. There's a 60% chance that we can make it totally miss the earth, a 20% chance China can make it mostly burn up and destroy a small frontier town in western China that can be easily evacuated, and a 20% chance its remains will sail over the Himalayas and crash within 50 miles of New Delhi. There's also a.5% chance we're totally wrong about its path, and if we are, it could crash into the South China Sea and wipe Hong Kong and the vicinity off the map. Do you let it (hopefully) crash harmlessly into the Gobi Desert and risk a catastrophe that's unlikely to happen, or do you take a much, much higher risk of causing lots of expensive damage (and probably deaths) by trying to screw with it in the hope that whatever you try won't end up making things worse?
As good as it feels to hate farm subsidies, they do serve a halfway useful purpose -- they basically eliminate famine and domestic shortage in exchange for higher total costs the other 95% of the time. Seriously... when's the last time you *ever* heard the word "famine" used in the context of "United States" or "Post-WWII non-Soviet-Bloc Europe"? If farmers operated purely without subsidy in a profit-maximizing way, they'd simply risk a bad food-free year every 10-20 years in exchange for.73% higher profits the next quarter. If one or two farmers did it, nobody would notice. If the American Agribusiness Industry acted like California's power-generation and transmission industry, we'd have a domestic crisis every time locusts descended upon Arkansas or Kansas (or at least poorer countries would, because the US would buy up most of their food).
Subsidizing dead industries is a bad thing, but there's a lot to be said for year-to-year stability as well. Would anybody who's sane *really* choose to save 1.9% per year in the long (25+ year horizon) run on groceries if it meant that prices at the store could soar overnight without warning, even if it meant that next year the same goods might be selling for pennies on the dollar? People have better things to do than spend their days researching prices and plan their purchase strategies for things they use daily at spot-market prices.
The reason for subsidies is simple -- it encourages farmers to plant enough to guarantee abundance under nearly any likely scenario, without leaving them trying to sell those same crops during a "good" year for less than they would have made by simply investing the season's crop capital in 6-month CDs and going on vacation somewhere. Gratuitous waste sucks, but shortage & famine is much, much worse.
That's just because the US and Canada have both perfected the fine art of maximally screwing each other's citizens and making cross-border commerce as painfully expensive as it can possibly be. It's not just Canadians who get raped buying stuff from America -- Americans buying stuff from Canada get raped every bit as badly. Hell, look at phone service... if there are two countries ANYWHERE on earth where it SHOULD be seamless, painless, and a complete non-issue to use a phone and data service on either side of the border, it's the US & Canada. Yet, data roaming charges if you're on the wrong side of the border relative to your country look like a nightmare out of 2002.
Hell, it was only about 5 years ago when Sprint & Telus customers got equally raped on the other side of their respective borders, despite the fact that back then, both networks were owned by Sprint, ran the same hardware, and probably shared a call center and back-end billing/provisioning infrastructure. But god help you if you had an unlocked Sprint/Telus phone, moved across the border, and wanted to activate it as a native phone on the "other" network. You could roam, but they'd never allow you to activate it for real even though the hardware was identical, and the firmware was literally a reflash away from being the same.
Aren't Australian TVs more expensive because Australia decided to go with a HDTV standard that was *just* different enough from the one used in the US and Europe to require specialization for sale in Australia? Like a modified form of 8-VSB modulation (used in the US), but 50fps frame/fieldrate (used in Europe and Australia), or something like that (basically, picking the worst parts of both American and European HD standards)?
Apple's prices might also be based upon the (now quite obsolete) assumption that AU$1 is nominally worth less than US$1. It's quite wrong now, but it wasn't very long ago that the exchange rate for YEARS was something like AU$1 ~= 50 US cents, instead of its current AU$1 ~= US$1.20. There's no excuse for that kind of mistake in hardware sales, but the old 2:1 exchange ratio was so well-established by history prior to a few years ago, it wouldn't surprise me if it weren't enshrined in music industry recording contracts that simply stated Australian royalties/licensing fees as being double the US amount or ~3-5 times the UK Pound amount.
Or, as others have speculated, Apple might just suck and be assholes because they can.
>. HP just did a half ass job of building market traction.
I'd say it's more like, they jumped the gun and tried too hard to monetize it too quickly. When the G1 came out, Google was smart. They created a phone that was a tiny bit underwhelming, but was mostly a fully-functional hardware reference platform that worked perfectly on at least one major US network, and worked equally well overseas. They didn't expect to ever make a cent from the G1's sales, because they understood that the G1's entire purpose was to *exist* and not flop. Period, full stop, end of story. Even then, Google's own mistake was obvious in retrospect -- they didn't accommodate AT&T's frequencies, and they didn't have a CDMA version. If you look at Android's growth, you have a series of small blips during the G1 era, then suddenly an overnight explosion in Q4 2009 when the Hero hit Sprint and the Droid hit Verizon. The point is, Google didn't even *try* to make real money on Android until it was well into its second generation of hardware. HP, meanwhile, had its executives chomping at the bit thinking they could roll out a fairly expensive device with minimal developer support and little in the way of real thirdparty software, and magically have it take off into a profitable project. When their delusional fantasy got shattered, they had nowhere left to go.
HP would have been FAR better off to have "softly" introduced WebOS as a product available for free for installation on select pre-existing Android hardware, and worried about making actual money from licensing fees on the NEXT round of hardware sold with it in the box. And to have changed the stupid name. I know it's an inaccurate perception, but even now, every time I hear the name "WebOS", I cringe & think of a lame half-assed web-based API built on Javascript that just kind of sucks the way 99.9% of browser-based apps do... but worse, because the underlying hardware itself is slower than the browser I've run sucky slow Javascript-based web apps in for comparison. It's a shame HP ran it into the ground, because it would have been good to have another good Unix-based mobile OS around to keep Android on its toes (I'll freely admit my prejudice, and state that I hate Apple & have looked down upon them since the day in middle school when I found out that the Mac didn't even have an apparent programming language AVAILABLE, let alone BASIC in the box. For the kids here, I believe that to actually write real software for a first-gen Mac, you actually had to buy a Lisa)
Your laptop's battery is larger than ~4cm x 4cm x 1/2cm? Because that's approximately the size of my Epic4G's 3500mAH extended battery. If your laptop is able to run with a battery that small, I'm impressed.
I think it's more like, Android achieved critical mass with software, so to prevail over Android at this point you'd have to not just be better -- you'd have to be "part the Red Sea and redefine the concept and very foundation of tablet-based computing" better. Or give up on trying to license WebOS as a commercial project, and work on getting Cyanogen et al to start porting it to popular Android phones as an alternative. But hey, four or five years ago, I was one of the ones screaming that Palm's only chance in hell to reboot its US developer ecosystem would have been to have ported it to the HTC PPC6700 and give it away along with their preview SDK (which would have instantly turned one of the most popular PDA phones of ~2006 into potential Palm Linux phones, and allowed Palm to bypass the slow wheels of change in Sprint/Verizon management and gotten them directly onto both networks through the back door). Obviously, that didn't happen, Palm further fucked up by tainting WebOS and supporting only lame web-based applications instead of making the whole SDK available from Day One, and pretty much finished it off by letting Sprint have an exclusive deal on WebOS phones for half a year while Europeans were begging for a GSM version, and hobbling their alleged flagship phone with subpar hardware to boot.
Palm's past ~6 year history is an epic study in how you can fuck up and destroy one of the most popular products in history, alienate 99% of your developers in the process (carriers didn't kill Cobalt... PODS did. Plain and simple, it sucked balls compared to CodeWarrior). HP was just the final chapter. Instead of trying to reinvent WebOS again, HP should have just cranked out a phone with the hardware specs of a Nexus One with a big 3500mAH battery and blown everyone away through brute force. Sexy consumer finesse can come next year. When you have almost no thirdparty software to speak of, priority one HAS to be jumpstarting your developer ecosystem. Developers don't care about thin and sexy, as long as the hardware is totally cool. Palm (and later, HP) got too caught up with premature optimization, and ended up missing the boat while Windows Mobile sailed off into the abyss and the Android Army arrived to take its place.
Yes, but they're *Apple* customers, so deep down inside, they know that Steve does it because he loves them and it's for their own good.
GM has customers too. Some are masochistic enough to buy GM cars over and over again just because They Believe In GM. Others feel compelled to buy only American cars, and show their disgust by buying a GM car, then a non-GM American car, then repeat the cycle. I use GM as an example, because they're the textbook example of a company that just plain doesn't "Get It", has never "Gotten It", and probably won't *ever* truly get it. The one division that showed empathy and hinted that it might have a clue (Saturn) was shut down because it was demoralizing the rest of the company. 50 years from now, Apple might very well be the next GM, living on its past glories and basking in its history.
As I understand it, the problem is that Samsung's kernel uses BlueZ as a module, but it was compiled without support for HID and SPP. To add HID and SPP, you'd have to recompile and rebuild the entire module, including the parts that interface with the largely undocumented and Samsung-proprietary Bluetooth chipset itself. In other words, you can't just build a module for HID and SPP and clamp it onto the existing module to extend its functionality. It's all or nothing -- total replacement, or nothing at all.
It would be pointless. Android is free and costs nothing to put on a phone. A phone without any OS whatsoever would be a warranty-claim nightmare, because you'd never be able to distinguish between a phone that doesn't work because of a hardware defect vs a phone that doesn't work because the user is an idiot. A stock firmware means you can take the phone that allegedly doesn't work, reflash it (if necessary) to that firmware, and declare the user to be an idiot with baseless claim if the phone works as advertised once reflashed.
It doesn't have to be black and white. By all means, ship the phones with stock firmware. Make it a tiny bit challenging to reflash, just to weed out the people who really don't have the necessary skills. Requiring some kind of physical hardware that's not exotic, but requires some kind of intentional act on the part of the user isn't a bad idea, either, just to make drive-by malware impossible (my personal vote goes to having the user plug in a wired headset and boot the phone while holding in the 'answer' button, then release it during a 2-3 second window of opportunity once a LED elsewhere on the phone blinks, then press some other combination of hardkeys just to make it clear that the user intends to launch the bootloader and it's not an accident). Hell, they can even go so far as to require cryptographically-signed updates, as long as the private signing key for MY phone is in MY possession so I can use it whenever I damn well feel like it.
>It stuns me that a rag-tag group of enthusiasts can so thoroughly spank a > billion dollar corporation's highly funded professional developer group.
A developer group that, like any, is thinly-spread across dozens of individual projects at any moment in time, and never has enough time to do more than make it work well enough to satisfy Marketing & Management. Companies like HTC and Samsung are starting to realize that a dozen guys porting CMx to their hardware is roughly equivalent to quadrupling the size of their in-house development staff. It's cheaper to hire a coach or two to manage an army of unpaid volunteers than it is to hire an equivalent number of full-time real employees.
As for forced obsolescence driving upgrades, let's be real -- these phones have an average best-case life expectancy of a year or two, max, before somebody drops them, smashes the screen, and ends up with a repair bill that exceeds the cost of a new phone if the user can manage to keep it from happening before his next upgrade is available. They don't HAVE to make them prematurely obsolete. Isaac Newton, slippery plastic, asphalt/concrete, and glass will do the job of removing them from use within a year or two anyway.
Phone manufacturers can be like GM, piss off & arbitrarily punish customers for the sin of owning last year's model, and guarantee that your next phone will be made by ANYBODY besides them, or they can be like Honda or Lexus, thoroughly *delight* customers with their phones up until the last day they own them, and practically assure that your next ten phones will be made by them as well. In this respect, Samsung kind of falls in the middle (like Hyundai) -- historically, not trying all that hard to delight, but not trying to *antagonize* customers (like GM) either. Like Hyundai, Samsung eventually seems to have realized the direction it wants to go in the future (delighted customers purchasing "affordable luxury" -- not really *cheap*, but shooting for a high feature-per-dollar ratio that puts them 98% of the way to the top for half the price of the next genuine step up)
Exactly. For the most part, Samsung has always been plagued by good (if not great) hardware, crippled by last year's software. Bringing Steve on board means they can now give him the keys to the candy store and let him directly play with the deepest secrets of Samsung's hardware without having to go all the way and make those same closely-guarded secrets public. Think about real-world Linux. How many Ubuntu, Debian, or Fedora users actually build their own kernels? Statistically, none. Between reasonably intelligent decisions at build-time, a half-dozen or so kernels optimized for mainstream bundles of real-world needs to choose from, and loadable kernel modules, there are few reasons for anyone not in academia to roll a kernel from scratch. With a little luck, Samsung users will BE in that position within a few months. Samsung has generally gotten the.ko LKM part right, they've just screwed up the rest of the Kernel. With Steve in there to ensure that there's a signed (if officially unsupported) alternate kernel or two to download from Samsung's website that does everything Cyanogen needs it to do, we'll be happy campers.
Steve, if you happen to read this, just one request for future Samsung kernels: Bluetooth HID and SPP. Please. You can even skip the downstream implementation for now. Just get the damn kernel-level stuff in there, so anybody with a rooted Samsung phone and stock Kondik-era kernel can take it from there and make it work later. It's the kernel that kills us dead in our tracks every single time, because rolling a custom kernel from scratch inevitably seems to mean giving up Sprint 4G, working network-accelerated GPS, or some other combination of basic hardware capabilities whose loss profoundly compromises the phone's worth and usability.
Now that Steve's with Samsung, I *might* actually consider buying an.*Epic.*Touch.*$ in 2 months if it ships with a dual-core processor and doesn't have some fatal, stupid last-minute cost-shaving hardware deficiency that ruins it just to reduce the manufacturing cost by 23 cents. The fact that Motorola might shift from Lawful Evil to Chaotic Neutral under Google definitely complicates matters a bit, since Tegra2 has Google's "Honeycomb Hometown" Advantage. Up to now, the Evo3D was a slam-dunk just because it was the most likely to end up with fully-functional Cyanogenmod. If every future Moto phone ends up being a de-facto Nexus-* and Samsung phones have Steve getting paid to make them the best-of-breed reference implementation for Cyanogen, HTC doesn't look quite as appealing.
Slashdot Mirror
^^^ That reminds me of another caveat -- inrush current. Powerful fans, in particular, are hard to use with generators -- even big ones. I have a Vornado fan (circa 1995) that can almost blow the bark off of a tree when it's running at full speed. My first generator (a 4-stroke 2000-watt baby generator like the one I described above) couldn't run it. I plugged it into the extension cord, turned it on (after starting the generator and letting it stabilize), and the generator literally rocked about 3 inches in the air on one side and choked to a halt as though an invisible hand just grabbed the spinning rotor and forced it to stop. The same generator was able to start a cheap window box fan... but ONLY if I quickly turned the knob from "off" to "medium" and allowed it to stabilize before turning it up to 'high'. If I went directly from "off" to "high", it would stall the generator.
The microwave oven was another thing that the generator didn't like *at all*. I tried using the microwave with generator #2 (5600-watt Craftsman). It worked, but both the microwave and generator made really bad-sounding noises (hard to describe, kind of a buzzing hum that was REALLY loud), and I decided to just forget about trying to use the microwave on generator power due to worries that it would damage the oven, the generator, or both.
That reminds me... if you're in the hurricane's path, do all your laundry now. You can run a washing machine from a generator, but even a whole-house 24-kW Generac is going to struggle with an electric dryer. I don't know about the northeast, but in Florida, clothes lines just don't work during the summer. You can leave clothes hanging on them all day, and they'll STILL be damp when the sun goes down. Post-Wilma, my coworkers and I had to bring damp clothes to the office and hang them on makeshift clotheslines between cubes to get them to dry out in the air conditioning.
After Hurricane Wilma, I had no power for almost four weeks.
Four. Fucking. Weeks.
I didn't live out in BFE, either... I lived in Coral Gables, which is about as hardcore "Central Dade County" as you can get.
That said, here's a big, huge tip for anybody who wants to be able to run a window air conditioner from a generator -- all things equal, the magic minimum is around 3,600 watts. I'd recommend 4200-4800 minimum. Why?
1. The generator's wattage is a polite fiction. The number printed on the box is roughly what it can output for about 5 minutes before Bad Things Happen. The REAL power output it can SUSTAIN is about 80% of that amount, maybe less.
2. Most generators have split-phase power, which is a nice way of saying that the big number printed by the box is kind of divided between two outlets. So your "4800-watt" generator is really more like 2400 watts (max) per outlet (which translates into about 1800 watts per outlet sustained). A small window air conditioner draws about 1200-1500 running watts, and needs about 1800 watts to start up.
Now, the half & half rule isn't quite set in stone... you can usually get away with drawing about 2400 watts (sustained) from a single outlet on a "4000-watt" generator with no load on the other outlet, but then you run into the next problem:
3. Generator run times are usually quoted at "50% load". If you have an air conditioner connected to one outlet of a 4KW generator, it's not really a "50% load", even if it's the only thing you're running. Why? Unbalanced loads make your fuel economy go WAY down. It won't quite suck down as much gas as a 100% load, but from my own experience, it'll act kind of like an 80-90% load fuel-wise. So if you're going to run a window air conditioner from a 4KW generator, you might as well plug the refrigerator (or another small air conditioner) into the other outlet and enjoy it, because at that point it will barely make a dent in your fuel use.
That said, don't go hog wild and buy a 10KW generator without a good reason. Especially not a cheap one. Most cheap generators do a really bad job of throttling down to accommodate reduced loads, and will burn almost as much gas with a nightlight as they will with a 50% load. It's a balancing act, and it's an important one, because if you're going to be feeding a generator for a few days, let alone a few weeks, a $40-50/day gas habit quickly becomes painful.
Oh, I almost forgot... there's one last catch...
4. Generators and UPSes don't get along. At all. 99.9% of the UPSes you can buy at a retail store will ignore electricity from a generator, will run 100% from the battery until it's drained, and shut down. There ARE expensive inverter-type generators that can charge a UPS, and UPSes that can charge from a cheap generator, but both are likely to cost more than it's worth spending.
4b. Generators and some DC power supplies don't get along very well, either. It's hit-or-miss, and hard to tell which power supplies are generator-unfriendly without testing them. Some will operate very, very inefficiently, and some won't work at all. The problem is that cheap (non inverter-type) generators don't output sine waves, and their "dirty" output doesn't play nicely with switching-type power supplies. You MIGHT be able to get around this by "double conversion". After Wilma, I had to power my DSL modem by plugging a 12v adapter into an outlet (which gave me a fake cigarette lighter rated at 1000mA), then plugged an inverter into it (giving me a 110v outlet), then plugged the DSL modem's power supply into the inverter. Ugly in countless ways, but it got me back online.
4c. As a corollary to 4b, most cheap generators suck at battery-charging.
The moral: if you don't need air conditioning, and can afford it, buy an inverter-type generator. They'll play nicely with power supplies (but your UPS might still get bitchy), and low-RPM expensive inverter-type generators also tend to be the quietest and most fuel efficient. Apparently, Honda makes some o
I meant, beyond the strait after it made landfall on the US side. Obviously the part under the strait itself is meant to be a tunnel. The specific obstacle I was referring to was not open water, but glacier fields where conventional civil engineering kind of falls flat on its face.
I'd be more concerned about permafrost, and possibly glaciers. Especially glaciers -- they're the one surface terrain where we're still pretty much in the stone age insofar as the construction of permanent precision infrastructure is concerned. You can't sink pilings down to the bedrock, because glaciers flow (slowly) and would eventually shear them away at "ground" level. You can't treat ice like bedrock, because pressure liquefies it and causes whatever you built on top to slowly sink. To cross a glacier field, you'd literally have to blast a canyon into the ice down to the permafrost, THEN build it on refrigerated concrete pilings to keep the ground around it frozen forever. Or maybe build a long multi-span suspension bridge whose support columns rested on concrete foundations big enough to flow the ice around them (with occasional seismic-like activity when the ice briefly prevails and manages to shove the concrete artificial island a few inches). Even if the environmentalists didn't go into convulsions, I'm sure it would be so cost-insane it would almost make sense to build the whole thing in a deep bored tunnel instead.
That said, if they can solve the civil engineering problems in a way that won't bankrupt the venture, it would be pretty cool. If nothing else, it would mean there would eventually be cheaper transportation links between Alaska and the lower 48 + Canada as well, so the state of Alaska would no longer have to budget enough food stamps for indigenous people who insist on living 500 miles from the nearest Wal-Mart and pay something like $12 for a box of Rice Krispies.
> A wave of locusts in Kansas would bump worldwide food costs up a couple of cents.
Assuming there's abundance elsewhere in the system to absorb the shock. The problem is, if you eliminate the subsidies and EVERYONE starts to farm for maximum shareholder value and return on investment, most of that worldwide capacity would eventually disappear and be wrung out of the system the same way the US mortgage industry was leveraged to the breaking point with toxic assets that COULDN'T be unloaded quickly enough to preserve necessary liquidity to avoid a collapse. It wouldn't happen in a year or two, but eventually, you WOULD end up in situations where a really bad crop failure in one or two parts of the US would be painful everywhere, because the other places that would have otherwise grown unprofitable (without subsidies) crops would be growing more profitable things instead (or be using the land to build suburban estate homes for people with horses, for example).
Growing lettuce in the Arizona desert isn't economically efficient, but it DOES help to guarantee that a major crop failure, um, wherever it is that lettuce normally gets grown will be reflected in a minor temporary price bump instead of lettuce simply becoming unavailable for weeks or months. Americans like knowing that we can go to the 'produce' section of the grocery store confident that we'll find exactly the same pristine, flawless fruits and vegetables we found last week, for more or less the same price, regardless of month, weather, or news reports. In other parts of the world, where things ARE left more up to pure market forces, that's NOT necessarily the case.
That's not to say the system is anywhere near perfect or entirely for the purpose of avoiding famine (tobacco being the obvious example). It does, however, go a long way towards making the availability of flawless fresh food something that "just works" in the background and most Americans never have to be bothered with.
China has an official government standard for micro-USB charging that basically says, "if the data lines are shorted together, the client can legitimately assume that the host is a charger capable of supplying 1.7 Amperes". By virtue of everything coming from China, that basically means that there's now a worldwide standard for the same. However, most cheap USB chargers DON'T short the pins together, because then they'd have to have a regulator with proper heatsink capable of supplying up to 1.7A continuously without melting or overheating. I believe most of the really, really cheap chargers use parts that will quickly die if you try to draw more than ~500mA continuously.
This is why, for example, a Samsung or HTC Android phone charges at least twice as fast when connected to their official chargers (and some thirdparty chargers) than it does when connected to a powered USB hub, and several orders of magnitude faster than they'll charge from a laptop's USB port. The bundled power supply shorts D+ and D-, so the phone knows it can draw 700mA (Samsung) or some comparable amount (HTC). Likewise, if the phone senses USB, but can't discern its power capabilities, it will assume the worst and limit its draw to 100mA (this is why quite a few Android phones will tread water instead of charge if you're tethered via USB... they're drawing power from the host PC, but less total current than they're using at that moment to maintain an active data connection.
Just wait until someone finally comes up with a viable standard for some hybrid between cellular and wi-fi (802.69?) that basically puts gigabit wiring back into the walls so it can feed hundreds of access points with service radii of ~10-25 feet apiece, and seamless hand-offs as users move around. Wi-fi is easy if your goal is to enable somebody with a laptop to get online at low speeds and not move around a lot. Enabling somebody to walk across the office and enjoy seamless wireless 100+mbit/sec connectivity every inch of the way is another matter entirely, and a problem companies like Cisco have barely even *started* to work on.
If you're going to go with football analogies, Apple is scoring a touchdown with two-point bonus every hour. Google, meanwhile, is massively spoiling their plans and running up the score by intercepting the ball every time Apple tries to do something and immediately kicking successful field goals from the 90-yard line. BCS is being grumpy and giving Apple lots of extra votes because it's scoring "legitimate" points while Google is gaming the system and scoring lots of points in ways BCS doesn't necessarily agree with, but at the end of the day, Google's going to go home victorious with a score of 90-24, even if Apple ends up getting invited to the best bowl game. Team Google doesn't care, because they won, stomped all over Apple, and they're getting laid tonight anyway after the victory party.
What you need is a USB CondomCable with the D+ and D- pins shorted together. No data can flow, and if the bad guys didn't bother to try and implement proper power protocol, you'll get the added satisfaction of frying THEIR hardware when your phone cranks up the juice and tries to suck down 1.7A instead of politely sipping 100mA. Just don't ever use such a cable by mistake to connect your phone to a pc or laptop belonging to yourself or a friend.
> I don't think we have the technology to print crap yet. I'm sure someone is working on it though.
Your wish is their command -- http://www.makerbot.com/ :-)
Frighteningly enough, I can see this happening. It would almost be better to plot the sail's trajectory to screw with its trajectory if it passes THROUGH the keyhole, and glide by harmlessly if Apophis MISSES the keyhole. By all means get the hardware into position to avert disaster... but if it looks like disaster isn't likely to happen, for the love of ${deity} don't go screwing with it and risk making things worse just for the sake of Doing Something.
It's kind of like theoretical weather-control experiments. If we someday know Tropical Storm Tyreesha is predicted to hit Miami on September 18th and disrupt the season finale of a popular mass-public reality TV show, do we REALLY want to risk kicking it back into the Straits of Florida and nudging it south so it passes harmlessly over the southern Everglades... then unexpectedly deviates from the models, grows into a monster category 5 hurricane over the Gulf, and destroys Houston the way Andrew destroyed Miami, instead of dumping a foot of rain on South Florida, fizzling out, sputtering back to life, and drifting north until it trips over southern Alabama as a minor category 1 hurricane?
Thought experiment: someday, a small asteroid is discovered that, if we do nothing, will strike a relatively desolate part of the earth and do about as much real damage as Tunguska. We have the ability to nudge it a bit. There's a 60% chance that we can make it totally miss the earth, a 20% chance China can make it mostly burn up and destroy a small frontier town in western China that can be easily evacuated, and a 20% chance its remains will sail over the Himalayas and crash within 50 miles of New Delhi. There's also a .5% chance we're totally wrong about its path, and if we are, it could crash into the South China Sea and wipe Hong Kong and the vicinity off the map. Do you let it (hopefully) crash harmlessly into the Gobi Desert and risk a catastrophe that's unlikely to happen, or do you take a much, much higher risk of causing lots of expensive damage (and probably deaths) by trying to screw with it in the hope that whatever you try won't end up making things worse?
As good as it feels to hate farm subsidies, they do serve a halfway useful purpose -- they basically eliminate famine and domestic shortage in exchange for higher total costs the other 95% of the time. Seriously... when's the last time you *ever* heard the word "famine" used in the context of "United States" or "Post-WWII non-Soviet-Bloc Europe"? If farmers operated purely without subsidy in a profit-maximizing way, they'd simply risk a bad food-free year every 10-20 years in exchange for .73% higher profits the next quarter. If one or two farmers did it, nobody would notice. If the American Agribusiness Industry acted like California's power-generation and transmission industry, we'd have a domestic crisis every time locusts descended upon Arkansas or Kansas (or at least poorer countries would, because the US would buy up most of their food).
Subsidizing dead industries is a bad thing, but there's a lot to be said for year-to-year stability as well. Would anybody who's sane *really* choose to save 1.9% per year in the long (25+ year horizon) run on groceries if it meant that prices at the store could soar overnight without warning, even if it meant that next year the same goods might be selling for pennies on the dollar? People have better things to do than spend their days researching prices and plan their purchase strategies for things they use daily at spot-market prices.
The reason for subsidies is simple -- it encourages farmers to plant enough to guarantee abundance under nearly any likely scenario, without leaving them trying to sell those same crops during a "good" year for less than they would have made by simply investing the season's crop capital in 6-month CDs and going on vacation somewhere. Gratuitous waste sucks, but shortage & famine is much, much worse.
That's just because the US and Canada have both perfected the fine art of maximally screwing each other's citizens and making cross-border commerce as painfully expensive as it can possibly be. It's not just Canadians who get raped buying stuff from America -- Americans buying stuff from Canada get raped every bit as badly. Hell, look at phone service... if there are two countries ANYWHERE on earth where it SHOULD be seamless, painless, and a complete non-issue to use a phone and data service on either side of the border, it's the US & Canada. Yet, data roaming charges if you're on the wrong side of the border relative to your country look like a nightmare out of 2002.
Hell, it was only about 5 years ago when Sprint & Telus customers got equally raped on the other side of their respective borders, despite the fact that back then, both networks were owned by Sprint, ran the same hardware, and probably shared a call center and back-end billing/provisioning infrastructure. But god help you if you had an unlocked Sprint/Telus phone, moved across the border, and wanted to activate it as a native phone on the "other" network. You could roam, but they'd never allow you to activate it for real even though the hardware was identical, and the firmware was literally a reflash away from being the same.
Aren't Australian TVs more expensive because Australia decided to go with a HDTV standard that was *just* different enough from the one used in the US and Europe to require specialization for sale in Australia? Like a modified form of 8-VSB modulation (used in the US), but 50fps frame/fieldrate (used in Europe and Australia), or something like that (basically, picking the worst parts of both American and European HD standards)?
Apple's prices might also be based upon the (now quite obsolete) assumption that AU$1 is nominally worth less than US$1. It's quite wrong now, but it wasn't very long ago that the exchange rate for YEARS was something like AU$1 ~= 50 US cents, instead of its current AU$1 ~= US$1.20. There's no excuse for that kind of mistake in hardware sales, but the old 2:1 exchange ratio was so well-established by history prior to a few years ago, it wouldn't surprise me if it weren't enshrined in music industry recording contracts that simply stated Australian royalties/licensing fees as being double the US amount or ~3-5 times the UK Pound amount.
Or, as others have speculated, Apple might just suck and be assholes because they can.
>. HP just did a half ass job of building market traction.
I'd say it's more like, they jumped the gun and tried too hard to monetize it too quickly. When the G1 came out, Google was smart. They created a phone that was a tiny bit underwhelming, but was mostly a fully-functional hardware reference platform that worked perfectly on at least one major US network, and worked equally well overseas. They didn't expect to ever make a cent from the G1's sales, because they understood that the G1's entire purpose was to *exist* and not flop. Period, full stop, end of story. Even then, Google's own mistake was obvious in retrospect -- they didn't accommodate AT&T's frequencies, and they didn't have a CDMA version. If you look at Android's growth, you have a series of small blips during the G1 era, then suddenly an overnight explosion in Q4 2009 when the Hero hit Sprint and the Droid hit Verizon. The point is, Google didn't even *try* to make real money on Android until it was well into its second generation of hardware. HP, meanwhile, had its executives chomping at the bit thinking they could roll out a fairly expensive device with minimal developer support and little in the way of real thirdparty software, and magically have it take off into a profitable project. When their delusional fantasy got shattered, they had nowhere left to go.
HP would have been FAR better off to have "softly" introduced WebOS as a product available for free for installation on select pre-existing Android hardware, and worried about making actual money from licensing fees on the NEXT round of hardware sold with it in the box. And to have changed the stupid name. I know it's an inaccurate perception, but even now, every time I hear the name "WebOS", I cringe & think of a lame half-assed web-based API built on Javascript that just kind of sucks the way 99.9% of browser-based apps do... but worse, because the underlying hardware itself is slower than the browser I've run sucky slow Javascript-based web apps in for comparison. It's a shame HP ran it into the ground, because it would have been good to have another good Unix-based mobile OS around to keep Android on its toes (I'll freely admit my prejudice, and state that I hate Apple & have looked down upon them since the day in middle school when I found out that the Mac didn't even have an apparent programming language AVAILABLE, let alone BASIC in the box. For the kids here, I believe that to actually write real software for a first-gen Mac, you actually had to buy a Lisa)
Your laptop's battery is larger than ~4cm x 4cm x 1/2cm? Because that's approximately the size of my Epic4G's 3500mAH extended battery. If your laptop is able to run with a battery that small, I'm impressed.
I think it's more like, Android achieved critical mass with software, so to prevail over Android at this point you'd have to not just be better -- you'd have to be "part the Red Sea and redefine the concept and very foundation of tablet-based computing" better. Or give up on trying to license WebOS as a commercial project, and work on getting Cyanogen et al to start porting it to popular Android phones as an alternative. But hey, four or five years ago, I was one of the ones screaming that Palm's only chance in hell to reboot its US developer ecosystem would have been to have ported it to the HTC PPC6700 and give it away along with their preview SDK (which would have instantly turned one of the most popular PDA phones of ~2006 into potential Palm Linux phones, and allowed Palm to bypass the slow wheels of change in Sprint/Verizon management and gotten them directly onto both networks through the back door). Obviously, that didn't happen, Palm further fucked up by tainting WebOS and supporting only lame web-based applications instead of making the whole SDK available from Day One, and pretty much finished it off by letting Sprint have an exclusive deal on WebOS phones for half a year while Europeans were begging for a GSM version, and hobbling their alleged flagship phone with subpar hardware to boot.
Palm's past ~6 year history is an epic study in how you can fuck up and destroy one of the most popular products in history, alienate 99% of your developers in the process (carriers didn't kill Cobalt... PODS did. Plain and simple, it sucked balls compared to CodeWarrior). HP was just the final chapter. Instead of trying to reinvent WebOS again, HP should have just cranked out a phone with the hardware specs of a Nexus One with a big 3500mAH battery and blown everyone away through brute force. Sexy consumer finesse can come next year. When you have almost no thirdparty software to speak of, priority one HAS to be jumpstarting your developer ecosystem. Developers don't care about thin and sexy, as long as the hardware is totally cool. Palm (and later, HP) got too caught up with premature optimization, and ended up missing the boat while Windows Mobile sailed off into the abyss and the Android Army arrived to take its place.
Yes, but they're *Apple* customers, so deep down inside, they know that Steve does it because he loves them and it's for their own good.
GM has customers too. Some are masochistic enough to buy GM cars over and over again just because They Believe In GM. Others feel compelled to buy only American cars, and show their disgust by buying a GM car, then a non-GM American car, then repeat the cycle. I use GM as an example, because they're the textbook example of a company that just plain doesn't "Get It", has never "Gotten It", and probably won't *ever* truly get it. The one division that showed empathy and hinted that it might have a clue (Saturn) was shut down because it was demoralizing the rest of the company. 50 years from now, Apple might very well be the next GM, living on its past glories and basking in its history.
As I understand it, the problem is that Samsung's kernel uses BlueZ as a module, but it was compiled without support for HID and SPP. To add HID and SPP, you'd have to recompile and rebuild the entire module, including the parts that interface with the largely undocumented and Samsung-proprietary Bluetooth chipset itself. In other words, you can't just build a module for HID and SPP and clamp it onto the existing module to extend its functionality. It's all or nothing -- total replacement, or nothing at all.
It would be pointless. Android is free and costs nothing to put on a phone. A phone without any OS whatsoever would be a warranty-claim nightmare, because you'd never be able to distinguish between a phone that doesn't work because of a hardware defect vs a phone that doesn't work because the user is an idiot. A stock firmware means you can take the phone that allegedly doesn't work, reflash it (if necessary) to that firmware, and declare the user to be an idiot with baseless claim if the phone works as advertised once reflashed.
It doesn't have to be black and white. By all means, ship the phones with stock firmware. Make it a tiny bit challenging to reflash, just to weed out the people who really don't have the necessary skills. Requiring some kind of physical hardware that's not exotic, but requires some kind of intentional act on the part of the user isn't a bad idea, either, just to make drive-by malware impossible (my personal vote goes to having the user plug in a wired headset and boot the phone while holding in the 'answer' button, then release it during a 2-3 second window of opportunity once a LED elsewhere on the phone blinks, then press some other combination of hardkeys just to make it clear that the user intends to launch the bootloader and it's not an accident). Hell, they can even go so far as to require cryptographically-signed updates, as long as the private signing key for MY phone is in MY possession so I can use it whenever I damn well feel like it.
>It stuns me that a rag-tag group of enthusiasts can so thoroughly spank a
> billion dollar corporation's highly funded professional developer group.
A developer group that, like any, is thinly-spread across dozens of individual projects at any moment in time, and never has enough time to do more than make it work well enough to satisfy Marketing & Management. Companies like HTC and Samsung are starting to realize that a dozen guys porting CMx to their hardware is roughly equivalent to quadrupling the size of their in-house development staff. It's cheaper to hire a coach or two to manage an army of unpaid volunteers than it is to hire an equivalent number of full-time real employees.
As for forced obsolescence driving upgrades, let's be real -- these phones have an average best-case life expectancy of a year or two, max, before somebody drops them, smashes the screen, and ends up with a repair bill that exceeds the cost of a new phone if the user can manage to keep it from happening before his next upgrade is available. They don't HAVE to make them prematurely obsolete. Isaac Newton, slippery plastic, asphalt/concrete, and glass will do the job of removing them from use within a year or two anyway.
Phone manufacturers can be like GM, piss off & arbitrarily punish customers for the sin of owning last year's model, and guarantee that your next phone will be made by ANYBODY besides them, or they can be like Honda or Lexus, thoroughly *delight* customers with their phones up until the last day they own them, and practically assure that your next ten phones will be made by them as well. In this respect, Samsung kind of falls in the middle (like Hyundai) -- historically, not trying all that hard to delight, but not trying to *antagonize* customers (like GM) either. Like Hyundai, Samsung eventually seems to have realized the direction it wants to go in the future (delighted customers purchasing "affordable luxury" -- not really *cheap*, but shooting for a high feature-per-dollar ratio that puts them 98% of the way to the top for half the price of the next genuine step up)
Exactly. For the most part, Samsung has always been plagued by good (if not great) hardware, crippled by last year's software. Bringing Steve on board means they can now give him the keys to the candy store and let him directly play with the deepest secrets of Samsung's hardware without having to go all the way and make those same closely-guarded secrets public. Think about real-world Linux. How many Ubuntu, Debian, or Fedora users actually build their own kernels? Statistically, none. Between reasonably intelligent decisions at build-time, a half-dozen or so kernels optimized for mainstream bundles of real-world needs to choose from, and loadable kernel modules, there are few reasons for anyone not in academia to roll a kernel from scratch. With a little luck, Samsung users will BE in that position within a few months. Samsung has generally gotten the .ko LKM part right, they've just screwed up the rest of the Kernel. With Steve in there to ensure that there's a signed (if officially unsupported) alternate kernel or two to download from Samsung's website that does everything Cyanogen needs it to do, we'll be happy campers.
Steve, if you happen to read this, just one request for future Samsung kernels: Bluetooth HID and SPP. Please. You can even skip the downstream implementation for now. Just get the damn kernel-level stuff in there, so anybody with a rooted Samsung phone and stock Kondik-era kernel can take it from there and make it work later. It's the kernel that kills us dead in our tracks every single time, because rolling a custom kernel from scratch inevitably seems to mean giving up Sprint 4G, working network-accelerated GPS, or some other combination of basic hardware capabilities whose loss profoundly compromises the phone's worth and usability.
Now that Steve's with Samsung, I *might* actually consider buying an .*Epic.*Touch.*$ in 2 months if it ships with a dual-core processor and doesn't have some fatal, stupid last-minute cost-shaving hardware deficiency that ruins it just to reduce the manufacturing cost by 23 cents. The fact that Motorola might shift from Lawful Evil to Chaotic Neutral under Google definitely complicates matters a bit, since Tegra2 has Google's "Honeycomb Hometown" Advantage. Up to now, the Evo3D was a slam-dunk just because it was the most likely to end up with fully-functional Cyanogenmod. If every future Moto phone ends up being a de-facto Nexus-* and Samsung phones have Steve getting paid to make them the best-of-breed reference implementation for Cyanogen, HTC doesn't look quite as appealing.
The day some nutcase PHB rams a horrible language like that down my throat is the day I officially begin development of a C++ to X-- compiler. ;-)