It goes against general operating practice and tradition, but strictly speaking, the only parties with the RIGHT to demand the source are parties who are licensed users of the GPL-licensed application itself. In other words, if HTC sells you an Android phone, you have the inalienable right to demand the GPL'ed source code. If I buy a Samsung Android phone, I have no right to demand that *HTC* furnish ME with a copy of the source, because I'm not their customer, and they never licensed Android to me. HTC can't stop my friend (who owns a HTC Android phone) from letting me have a copy, but they themselves have no affirmative duty to lift a finger and make it available to *me*.
Where things get ugly is the fact that HTC traditionally compiles its kernels into a monolithic binary blob, so the source is basically useless if your end goal is a working, fully-functional kernel suitable for a newer version of Android. Samsung, in contrast, DOES keep their proprietary modules neatly separated out as proper loadable kernel modules. A year ago, I thought that was a really big deal, and made Samsung heroes. Unfortunately, back then, I had no idea that Linux doesn't have a stable ABI, and a kernel module built for 2.6.32 is likely to be useless with a 2.6.35 kernel. Sigh. So, it looks like my next phone will be HTC, because (unlike Samsung), at least HTC tends to "unofficially" release new kernel builds in a timely manner instead of waiting for hell to freeze over. Now, if only they could be bothered to compile them with BlueZ HID profile enabled...
In the real world, here's what happens to pop stars: their song becomes #1, they sing and perform it a bazillion times, get burned out of singing and performing that song and vow to never, ever sing it again, and literally go for 20 years without so much as humming the melody. Then their kids get to college, and Princeton/Yale/Harvard is pretty expensive, even if you ARE fairly wealthy. They also start feeling old, and decide to go on tour to rekindle the flame, relive their youth, and pay their kids' tuition -- and at that point, they literally haven't sung the song in 20 years. If they're lucky, they kept singing for fun in the meantime, and still have that skill & their voices, so they can relearn it. If not... well... they get to lipsync a lot and disappoint people at dog tracks, fairs, and college homecoming events for the next couple of years.
The point is, 20+ years down the road, quite a few former popstars would be better off NOT trying to resurrect their careers by performing their old songs. A few artists pull it off and shine. The other 97% end up disillusioning and disappointing their fans.
Er... I think the Evo 4G just about the most thoroughly ROM-hacked phone since the PPC-6700 (the phone that basically put XDAdevelopers.com on the map). I know the Evo4G got off to a rocky start for the first month or two, but I'm pretty sure it had custom kernels and more Android distros than you could count without getting bored by mid-December.
Or are you talking about the Evo3G, which is a newer phone despite having a smaller number (and in any case, isn't a Motorola phone).
It's cultural. Motorola has a long history as a military defense contractor. They really, truly understand hardware encryption, and they use it with (almost) everything they touch simply because they *can* (and partly to use things like phones as a proving ground, showroom, and laboratory for new security technologies they have at their disposal). What better sales pitch for Motorola to throw at Sony and/or Microsoft when trying to sell security chips to lock down their next consumer game console than the fact that the bootloaders on the DroidX/Droid3/etc are still uncracked despite some very smart, very motivating people trying *really hard* for a year or more?
The more secure you make an embedded device or appliance against information leakage and harvesting-type vulnerabilities, the more likely it is to end up getting returned to stores by frustrated consumers who can't get it to work.
Just look at WPA-2 -- it's unquestionably more secure than WEP. It's also rarely used in public settings because statistically, it never fucking works. You can take any access point, and any device that supposedly supports WPA-2, and know beyond doubt that there's about a 50-50 chance it won't work on the first try, and only slightly better odds that it'll eventually work after an hour or more of work (likely victims include anything with Vista or newer, or an Android phone that hasn't been rooted & reflashed to AOSP or Cyanogen.
>But I don't see ATSC being replaced with yet another standard that will require yet another >converter box in the near future.
No, it's simply become as irrelevant to 98% of Americans as VHF and UHF NTSC tuners were 15 years ago (when just about everyone had a set top box from the cable or satellite company that did the job the useless, unloved, rarely-bothered-with tuner inside the TV was allegedly intended to do).
ATSC *almost* became relevant when DirecTV prepared to roll out its home central tuner system that would tune channels and remodulate them using Zenith's designed-but-never-sold 8vsb home modulator chipset, but Hollywood sent out its attack lawyers to strangle it in its cradle.
You're right in the sense that if Apple & Microsoft were able to banish new Android phones from America, a large group would jailbreak their phones and reflash them to Android in pure spite and disgust, but make no mistake... Android is as good as it is largely because well-funded companies pay lots of developers to work on it full time. If that development dried up, Android would end up kind of like Netbeans -- still still viable, but nowhere near as vibrant, polished, and cutting-edge as it was pre-Oracle.
Before anybody mentions Cyanogen, try finding a non-HTC phone running Cyanogen that has working Sprint 4G and server-accelerated GPS. Last I checked, reflashing an Epic 4G to CM6 or CM7 still means giving up wimax (prove me wrong, and I'll be delighted, because that's the only reason I haven't switched... but it's a big, important reason). Now, imagine the plight of Sprint Samsung phone owners trying to make Cyanogen work with wimax, and extend those same kind of problems to every new incremental technology improvement. If HTC and Samsung couldn't sell Android phones that work on Sprint and Verizon, Android phones on Sprint & Verizon would quickly become dysfunctional. Before GSM users gloat about CDMA users, go double-check and see just how many imported UMTS phones can do quasi-4G on T-Mobile, or get the highest available speeds (where available) on AT&T's variant of LTE. Hell, it's hard enough to find an imported phone not officially sold in the US that can even do fsck'ing *UMTS* 3G on American frequencies, let alone do it as quickly as AT&T and T-Mobile's newest-gen phones.
Make no mistake -- in America, if Android gets marginalized and kicked underground, life as an Android user on an American user will quickly start to suck within a year or two as new improvements appear that don't quite work right (or work at all) under Android.
> The ONLY things that should be patentable are physical devices. > ANYTHING other then that is covered by copyright.
Actually, that argument is kind of like advocating cutting off one's nose to spite one's face. As bad as software patents are, replacing them with copyrights would be worse. At least patents aren't eternal in duration and expire someday. If only literal sourcecode could be copyrighted, I'd be in agreement. However, if you're going to allow copyright to be blurred into 'look & feel' issues that currently get rammed into the 'patent' realm, it would quickly become impossible to write any kind of meaningful program for commercial gain.
> The idea that I can "create" something intangible, easily replicated, and quite literally out of nothing simply by typing some characters on > a keyboard is absolutely insane, and should never have been allowed in the first place.
Believe it or not, it gets worse. About 10-15 years ago, copyright was extended beyond mere mechanical reproduction of blueprints to actual architectural design. So, someone who comes up with an "innovation" like a garage with built-in recess to accommodate an electric-car charger could conceivably sue anybody else who designs/builds a house with a similar charger alcove. Had similar protection existed 50 years ago, things like attached 3-car garages with larger third bay for a boat or RV, pass-through breakfast bars, and bedrooms with adjacent-but-separate sitting areas could have been copyrighted for 100+ years as well. So far, most of the lawsuits have been over things like exterior appearance and between builders copying each other's floorplans, but the law itself imposes no real hard limits to what can be considered copyrightable, nor any minimum threshold before something can be classified as "infringing".
>Did it take two people 20 years to build ? No ? Did it take 20 people two years to build ?
I'm not sure about YOUR part of the country, but in places like Florida, California, and New York, land accounts for at least half the cost of a single-family home. When you consider that almost any multi-family building is going to end up on land that had something else on it first (and had to be purchased at a price that covered whatever was already there, then demolished and cleared away), land costs for multi-family buildings per-unit aren't a whole lot better.
It also assumes that most organizations actually CONFIGURE their servers to require such a passphrase on reboot, instead of storing the key in a readonly file owned by root to avoid the possibility of a web server that's down for any extended length of time because one of the people who know the passphrase weren't able to get to it immediately and ENTER that passphrase when something caused a reboot. You can pretty much take as a given that any entity smaller than a bank with 24/7 IT staff monitoring the server around the clock is going to do the safe (from a downtime perspective) thing and store the key in a file so the server can restart as quickly as possible after an unexpected reboot.
> OTOH, how do you export the services of a hairdresser?
It might not be possible to export hairdressing services, but it's most certainly possible to export services like "answer the telephone and do customer service for the customers of some company living in a country 6,000 miles away on the other side of the planet and (in theory) speak the same language you learned in high school" (ie, Indian call centers).
Ditto, for things like database administration. There's absolutely *nothing* to stop a company from Florida from leasing servers and rackspace in London that are 99.999% administered off-site by DBAs in Bangalore. It is, however, significantly cheaper to locate the servers themselves in the US or Europe. Why? The one thing that saves US & European competitiveness in data centers is the fact that it's fairly cheap and easy to get multiple levels of redundant communications and power infrastructure in place. And this partly goes to show why having good infrastructure is so important, even if it's built with deficit spending. It also illustrates why I've grudgingly come to see why allowing profit-maximization in things like power generation isn't necessarily a great idea. It might be cost-effective to bleat about "smart grids" and accepting a few hours of rolling blackouts per year in exchange for 3% higher quarterly profits by eliminating excess peak capacity, but those same decisions can disqualify entire regions from consideration by companies looking for reliable power. Rolling blackouts are something that happens in fucking second-rate countries, and the fact that there are people who'd willingly allow them to happen by design in the US as a matter of cost-efficiency should have elected officials getting hung out to dry on election day.
> and really switching from one database to another is in most cases > not entirely painful on the software side.
Um, I can assure you that porting an application that uses queries that were highly-optimized for MySQL performance is very, very, very painful. The streets of Silicon Valley are littered with the corpses of web-based startups that grew explosively, tried moving from MySQL to Oracle, and died along the way.
Re:Joe Sixpack isn't even using his 1080p right
on
Beyond HDTV
·
· Score: 1
Not quite... Nyquist is a *minimum*, not a *guarantee*. Nyquist only guarantees that a sampling rate less than double the highest-frequency component will suck and have artifacts.
Take 24fps source. Watch it on a 60fps progressive display by displaying alternating frames in 2-3-2-3 cadence. Now watch the same 24fps source on a progressive display with 120fps native framerate by showing each frame 5 times in a row. You most certainly WILL see a difference between the two, even though Nyquist (as commonly misunderstood by just about everyone) seems to imply that 2-3 vs 5-5 doesn't matter since a 60hz refresh rate is more than double the 24fps source rate.
Don't feel bad. It's a mistake almost everyone makes until someone points it out to them, because over the past 25 years magazines and books have tried to dumb down the definition of Nyquist to make it more understandable to the masses, but they've dumbed down the definition in ways that end up being wrong if you try pushing it to limits that were only theoretical goalposts 25 years ago.
As a practical matter, if you're going 70mph on a freeway that's covered in stormwater, you're going to be hydroplaning more or less nonstop whether you want to admit it or not. The difference is that a truck/SUV with 18" or larger tires is going to have deeper treads and greater water-shedding capacity, so it will hydroplane a tiny bit less and be less likely to get into a wreck than a smaller car (and more likely to survive the wreck if it happens). Road conditions that would be utterly suicidal for somebody in a tiny SmartCar get yawned at by people with big SUVs and trucks that are technically (if they weren't too expensive to risk scratching the paint) offroad-capable.
Re:Joe Sixpack isn't even using his 1080p right
on
Beyond HDTV
·
· Score: 1
"VGA", as in the official 640x480 video standard created by IBM in the mid-80s, has always -- by definition -- been a progressive-scan 60fps video mode. There was a weird video mode with an IBM-ish name something like "8514" that I believe was 1024x768, but 640x480 has never, ever been interlaced in any mainstream scenario on any mainstream PC videocard or monitor bearing the name "VGA". The only mainstream computer with an interlaced 640x480(-ish) video mode was the Amiga, and its 640x400 interlaced video mode was generally regarded as unusable by pretty much everyone who didn't have a FlickerFixer card to deinterlace it and make it useful.
Higher framerate. Please. I've seen glass-smooth 720p120 CGI on high-end computer monitors. It's so "lifelike" it's creepy. When you get to 100+ fps, video almost starts to look more "real" than real life (partly because we're all so conditioned from a lifetime of 24, 50, and 60fps video, anything faster automatically triggers the "must be real" visual reflex).
And no, I'm not talking about "100hz", "120hz", "200hz", or "240hz" TVs. They aren't increasing the framerate -- they're just using the faster refresh rate with oversampling to simulate interlace blur and scanline fade (partly because high-quality motion-vector deinterlacing is hard to do well, and pretty much impossible to do in realtime, so the only alternative to 1080i60 video with bob and weave artifacts on inherently-progressive displays is to simulate the way CRT scanlines used to blur and fade.
And yes, I'm one of those weird people who likes to rip 1080i60 interlaced content, then let something like Avisynth with MVtools chew on it for a week to decompose each scene into a panning/zooming "background" with moving "sprites", then tween them and synthesize missing detail to make everything look like real 60fps progressive-scan live video.:-)
Re:Joe Sixpack isn't even using his 1080p right
on
Beyond HDTV
·
· Score: 2
>It's all advertising.
Sorry, but no. No. A thousand times, "No!"
You can tell the difference between 480i and *anything* progressive on a 19" display across a smoky room the day after having laser eye surgery. It's the same difference you saw years ago when comparing any TV to any VGA monitor. It might have only been 640x480, but it had a pleasing "solid" appearance that was instantly visible compared to any TV.
That said, the difference between 480p60 and higher-res isn't quite as dramatic. 720p60 is a nice step up, but it's not night-and-day. 1080i60 can be better if there's not a lot of motion (to cause weave artifacts), but most of the time it gets bobbed into something with 960-1440 real horizontal pixels and 540 vertical pixels anyway. True 1080p60 is a sight to behold, but so little non-CGI real honest-to-god 1080p60 content exists, most of the time you're just seeing 1080p24 with every other frame shown 3 times in a row instead of 2, or maybe 1080i60 that's been deinterlaced with studio-grade postproduction hardware.
Also, "normal viewing distance" is so last-century. Real people with 60+ inch TVs sit almost close enough to touch the screen, because there's no *reason* to sit farther -- no screen door effect, no x-rays, and some nice perspective-filling immersion when you sit close to the screen.
> I was specifically commenting on your assertion that the market was demanding big cars, > when in fact that market evaporated even before the economic crises.
Unless you count export markets for American cars. When somebody in Japan, Australia, Britain, Saudi Arabia, Russia, Mexico, or elsewhere in the world buys an imported American car, it's NOT going to be an American-made Ford Focus or Chevy Aveo... it's going to be a Mustang, a Corvette, an Escalade, or maybe even a big F-150 truck. *THOSE* are the American cars with viable export markets. If they were eliminated, America would basically cease to have cars anybody in a foreign country would be interested in buying. They have plenty of domestic econoboxes to choose from.
American automakers know what their niche market is, and it's a niche they understand well. Pre-bankruptcy, Chrysler developed a very cool hybrid turbo-electric drivetrain that used the electric traction motor with PID feedback to compensate for turbo lag. Slam the pedal to the floor, and the electric motor kicks in to add a few dozen horsepower while the turbine is spinning up. The net result is a sports car that gets ~20-24mpg instead of the usual 12-18mpg, and basically lets you have your cake & eat it too.
> the vast majority of SUV owners do not haul cargo, do not drive off-road, and do not frequently carry more than 3 passengers.
Apparently, you don't live in South Florida and drive home from work on freeways that have visible sheets of water, and neighborhood streets under INCHES of water after any meaningful rainstorm. This is our daily reality.
In Miami, a big SUV or pickup truck means the difference between being stranded for hours wherever you were when it started raining, and being able to hydroplane home at 70mph on the freeway in some degree of relative safety. As an individual, there's not much I can do to unclog the dysfunctional storm drains of Dade & Broward counties, and even less I can do to make it stop raining for days at a time. However, I can personally go out and buy a vehicle capable of driving through 8" of water without major risk to the engine block and mitigate 90% of the problems that would otherwise accompany months of endless rainstorms.
Big, huge bit of advice: don't link your Market developer account to a Gmail account that does anything besides sit passively and be an otherwise-unused Gmail account. If you associated your Market developer account with a gmail account associated with G+ and they suspended it, you'd be *beyond* screwed commercially. Depending upon how aggressively Google wanted to mess up, you might not even be allowed to *create* a new developer's account for Android Market (it would see your credit card number, do a credit check lookup to verify your identity, realize it was associated with a suspended developer's account, and most likely either refuse to allow its creation, or nuke it without refund within hours or days.
THAT's what people are screaming about -- the fact that Google is basically setting people up for domino-like failure cascades that could have devastating professional and commercial consequences, and there are more than a few scenarios where they could happen to somebody who's either outright innocent, or at least didn't deserve the equivalent of an online death penalty.
Well... if Google effectively revoked the purchase of everything I'd purchased through Android Market without refund or recourse by suspending my Gmail account (the only way to purchase anything through Android Market), I'd be pretty pissed and would probably sue them just as a matter of principle. It's one thing to take away somebody's email service. It's another thing to retroactively confiscate purchased software whose only available means of mainstream purchase is with that Gmail account (or, in the case of AT&T, the only way to purchase apps, period... at least for now).
Even if the apps remained working until the phone were reflashed or replaced, I reflash my phone frequently as is my inalienable Linus/RMS-given right. If Google wants to be able to kill Gmail accounts for any reason (or no reason at all), they'd better damn well come up with a parallel SSO identifier for Android phones that can survive Google Exile unscathed. ESPECIALLY given their "shoot first, then be disinterested in even pretending to care about asking questions later" policy. At this rate, by the end of the year, there will be at least a dozen movie plots circulating around Hollywood for a proposed movie called "GoogleX'ed" whose plot is somebody who gets exiled from Google and has his life cascade and snowball into disintegration around him as service after service gets cancelled because it depends upon some other service that ultimately was tied to a Google service.
Er, well, if you want to talk about "stereo" gear sold at K-mart... maybe. But if you stepped up even slightly (say, to Service Merchandise and Circuit City, late 80s) and considered Technics to be the low end baseline, things were much better.
Plus, let's not forget the coffee table sized monster subwoofers DAK used to sell and ship by motor freight. I still have one upstairs in the bonus room;-)
Slashdot Mirror
It goes against general operating practice and tradition, but strictly speaking, the only parties with the RIGHT to demand the source are parties who are licensed users of the GPL-licensed application itself. In other words, if HTC sells you an Android phone, you have the inalienable right to demand the GPL'ed source code. If I buy a Samsung Android phone, I have no right to demand that *HTC* furnish ME with a copy of the source, because I'm not their customer, and they never licensed Android to me. HTC can't stop my friend (who owns a HTC Android phone) from letting me have a copy, but they themselves have no affirmative duty to lift a finger and make it available to *me*.
Where things get ugly is the fact that HTC traditionally compiles its kernels into a monolithic binary blob, so the source is basically useless if your end goal is a working, fully-functional kernel suitable for a newer version of Android. Samsung, in contrast, DOES keep their proprietary modules neatly separated out as proper loadable kernel modules. A year ago, I thought that was a really big deal, and made Samsung heroes. Unfortunately, back then, I had no idea that Linux doesn't have a stable ABI, and a kernel module built for 2.6.32 is likely to be useless with a 2.6.35 kernel. Sigh. So, it looks like my next phone will be HTC, because (unlike Samsung), at least HTC tends to "unofficially" release new kernel builds in a timely manner instead of waiting for hell to freeze over. Now, if only they could be bothered to compile them with BlueZ HID profile enabled...
In the real world, here's what happens to pop stars: their song becomes #1, they sing and perform it a bazillion times, get burned out of singing and performing that song and vow to never, ever sing it again, and literally go for 20 years without so much as humming the melody. Then their kids get to college, and Princeton/Yale/Harvard is pretty expensive, even if you ARE fairly wealthy. They also start feeling old, and decide to go on tour to rekindle the flame, relive their youth, and pay their kids' tuition -- and at that point, they literally haven't sung the song in 20 years. If they're lucky, they kept singing for fun in the meantime, and still have that skill & their voices, so they can relearn it. If not... well... they get to lipsync a lot and disappoint people at dog tracks, fairs, and college homecoming events for the next couple of years.
The point is, 20+ years down the road, quite a few former popstars would be better off NOT trying to resurrect their careers by performing their old songs. A few artists pull it off and shine. The other 97% end up disillusioning and disappointing their fans.
> Hopefully, they will unlock the Evo 4g
Er... I think the Evo 4G just about the most thoroughly ROM-hacked phone since the PPC-6700 (the phone that basically put XDAdevelopers.com on the map). I know the Evo4G got off to a rocky start for the first month or two, but I'm pretty sure it had custom kernels and more Android distros than you could count without getting bored by mid-December.
Or are you talking about the Evo3G, which is a newer phone despite having a smaller number (and in any case, isn't a Motorola phone).
It's cultural. Motorola has a long history as a military defense contractor. They really, truly understand hardware encryption, and they use it with (almost) everything they touch simply because they *can* (and partly to use things like phones as a proving ground, showroom, and laboratory for new security technologies they have at their disposal). What better sales pitch for Motorola to throw at Sony and/or Microsoft when trying to sell security chips to lock down their next consumer game console than the fact that the bootloaders on the DroidX/Droid3/etc are still uncracked despite some very smart, very motivating people trying *really hard* for a year or more?
The more secure you make an embedded device or appliance against information leakage and harvesting-type vulnerabilities, the more likely it is to end up getting returned to stores by frustrated consumers who can't get it to work.
Just look at WPA-2 -- it's unquestionably more secure than WEP. It's also rarely used in public settings because statistically, it never fucking works. You can take any access point, and any device that supposedly supports WPA-2, and know beyond doubt that there's about a 50-50 chance it won't work on the first try, and only slightly better odds that it'll eventually work after an hour or more of work (likely victims include anything with Vista or newer, or an Android phone that hasn't been rooted & reflashed to AOSP or Cyanogen.
>But I don't see ATSC being replaced with yet another standard that will require yet another
>converter box in the near future.
No, it's simply become as irrelevant to 98% of Americans as VHF and UHF NTSC tuners were 15 years ago (when just about everyone had a set top box from the cable or satellite company that did the job the useless, unloved, rarely-bothered-with tuner inside the TV was allegedly intended to do).
ATSC *almost* became relevant when DirecTV prepared to roll out its home central tuner system that would tune channels and remodulate them using Zenith's designed-but-never-sold 8vsb home modulator chipset, but Hollywood sent out its attack lawyers to strangle it in its cradle.
>What possible reason would you buy this over an iPad, other than "OMG I HATES THE APPLE"?
Freedom to install software not blessed by Saint Steven d'Cupertino, directly from .apk if you feel like it?
You're right in the sense that if Apple & Microsoft were able to banish new Android phones from America, a large group would jailbreak their phones and reflash them to Android in pure spite and disgust, but make no mistake... Android is as good as it is largely because well-funded companies pay lots of developers to work on it full time. If that development dried up, Android would end up kind of like Netbeans -- still still viable, but nowhere near as vibrant, polished, and cutting-edge as it was pre-Oracle.
Before anybody mentions Cyanogen, try finding a non-HTC phone running Cyanogen that has working Sprint 4G and server-accelerated GPS. Last I checked, reflashing an Epic 4G to CM6 or CM7 still means giving up wimax (prove me wrong, and I'll be delighted, because that's the only reason I haven't switched... but it's a big, important reason). Now, imagine the plight of Sprint Samsung phone owners trying to make Cyanogen work with wimax, and extend those same kind of problems to every new incremental technology improvement. If HTC and Samsung couldn't sell Android phones that work on Sprint and Verizon, Android phones on Sprint & Verizon would quickly become dysfunctional. Before GSM users gloat about CDMA users, go double-check and see just how many imported UMTS phones can do quasi-4G on T-Mobile, or get the highest available speeds (where available) on AT&T's variant of LTE. Hell, it's hard enough to find an imported phone not officially sold in the US that can even do fsck'ing *UMTS* 3G on American frequencies, let alone do it as quickly as AT&T and T-Mobile's newest-gen phones.
Make no mistake -- in America, if Android gets marginalized and kicked underground, life as an Android user on an American user will quickly start to suck within a year or two as new improvements appear that don't quite work right (or work at all) under Android.
> The ONLY things that should be patentable are physical devices.
> ANYTHING other then that is covered by copyright.
Actually, that argument is kind of like advocating cutting off one's nose to spite one's face. As bad as software patents are, replacing them with copyrights would be worse. At least patents aren't eternal in duration and expire someday. If only literal sourcecode could be copyrighted, I'd be in agreement. However, if you're going to allow copyright to be blurred into 'look & feel' issues that currently get rammed into the 'patent' realm, it would quickly become impossible to write any kind of meaningful program for commercial gain.
> The idea that I can "create" something intangible, easily replicated, and quite literally out of nothing simply by typing some characters on
> a keyboard is absolutely insane, and should never have been allowed in the first place.
Believe it or not, it gets worse. About 10-15 years ago, copyright was extended beyond mere mechanical reproduction of blueprints to actual architectural design. So, someone who comes up with an "innovation" like a garage with built-in recess to accommodate an electric-car charger could conceivably sue anybody else who designs/builds a house with a similar charger alcove. Had similar protection existed 50 years ago, things like attached 3-car garages with larger third bay for a boat or RV, pass-through breakfast bars, and bedrooms with adjacent-but-separate sitting areas could have been copyrighted for 100+ years as well. So far, most of the lawsuits have been over things like exterior appearance and between builders copying each other's floorplans, but the law itself imposes no real hard limits to what can be considered copyrightable, nor any minimum threshold before something can be classified as "infringing".
>Did it take two people 20 years to build ? No ? Did it take 20 people two years to build ?
I'm not sure about YOUR part of the country, but in places like Florida, California, and New York, land accounts for at least half the cost of a single-family home. When you consider that almost any multi-family building is going to end up on land that had something else on it first (and had to be purchased at a price that covered whatever was already there, then demolished and cleared away), land costs for multi-family buildings per-unit aren't a whole lot better.
It also assumes that most organizations actually CONFIGURE their servers to require such a passphrase on reboot, instead of storing the key in a readonly file owned by root to avoid the possibility of a web server that's down for any extended length of time because one of the people who know the passphrase weren't able to get to it immediately and ENTER that passphrase when something caused a reboot. You can pretty much take as a given that any entity smaller than a bank with 24/7 IT staff monitoring the server around the clock is going to do the safe (from a downtime perspective) thing and store the key in a file so the server can restart as quickly as possible after an unexpected reboot.
> OTOH, how do you export the services of a hairdresser?
It might not be possible to export hairdressing services, but it's most certainly possible to export services like "answer the telephone and do customer service for the customers of some company living in a country 6,000 miles away on the other side of the planet and (in theory) speak the same language you learned in high school" (ie, Indian call centers).
Ditto, for things like database administration. There's absolutely *nothing* to stop a company from Florida from leasing servers and rackspace in London that are 99.999% administered off-site by DBAs in Bangalore. It is, however, significantly cheaper to locate the servers themselves in the US or Europe. Why? The one thing that saves US & European competitiveness in data centers is the fact that it's fairly cheap and easy to get multiple levels of redundant communications and power infrastructure in place. And this partly goes to show why having good infrastructure is so important, even if it's built with deficit spending. It also illustrates why I've grudgingly come to see why allowing profit-maximization in things like power generation isn't necessarily a great idea. It might be cost-effective to bleat about "smart grids" and accepting a few hours of rolling blackouts per year in exchange for 3% higher quarterly profits by eliminating excess peak capacity, but those same decisions can disqualify entire regions from consideration by companies looking for reliable power. Rolling blackouts are something that happens in fucking second-rate countries, and the fact that there are people who'd willingly allow them to happen by design in the US as a matter of cost-efficiency should have elected officials getting hung out to dry on election day.
> and really switching from one database to another is in most cases
> not entirely painful on the software side.
Um, I can assure you that porting an application that uses queries that were highly-optimized for MySQL performance is very, very, very painful. The streets of Silicon Valley are littered with the corpses of web-based startups that grew explosively, tried moving from MySQL to Oracle, and died along the way.
Not quite... Nyquist is a *minimum*, not a *guarantee*. Nyquist only guarantees that a sampling rate less than double the highest-frequency component will suck and have artifacts.
Take 24fps source. Watch it on a 60fps progressive display by displaying alternating frames in 2-3-2-3 cadence. Now watch the same 24fps source on a progressive display with 120fps native framerate by showing each frame 5 times in a row. You most certainly WILL see a difference between the two, even though Nyquist (as commonly misunderstood by just about everyone) seems to imply that 2-3 vs 5-5 doesn't matter since a 60hz refresh rate is more than double the 24fps source rate.
Don't feel bad. It's a mistake almost everyone makes until someone points it out to them, because over the past 25 years magazines and books have tried to dumb down the definition of Nyquist to make it more understandable to the masses, but they've dumbed down the definition in ways that end up being wrong if you try pushing it to limits that were only theoretical goalposts 25 years ago.
As a practical matter, if you're going 70mph on a freeway that's covered in stormwater, you're going to be hydroplaning more or less nonstop whether you want to admit it or not. The difference is that a truck/SUV with 18" or larger tires is going to have deeper treads and greater water-shedding capacity, so it will hydroplane a tiny bit less and be less likely to get into a wreck than a smaller car (and more likely to survive the wreck if it happens). Road conditions that would be utterly suicidal for somebody in a tiny SmartCar get yawned at by people with big SUVs and trucks that are technically (if they weren't too expensive to risk scratching the paint) offroad-capable.
"VGA", as in the official 640x480 video standard created by IBM in the mid-80s, has always -- by definition -- been a progressive-scan 60fps video mode. There was a weird video mode with an IBM-ish name something like "8514" that I believe was 1024x768, but 640x480 has never, ever been interlaced in any mainstream scenario on any mainstream PC videocard or monitor bearing the name "VGA". The only mainstream computer with an interlaced 640x480(-ish) video mode was the Amiga, and its 640x400 interlaced video mode was generally regarded as unusable by pretty much everyone who didn't have a FlickerFixer card to deinterlace it and make it useful.
Higher framerate. Please. I've seen glass-smooth 720p120 CGI on high-end computer monitors. It's so "lifelike" it's creepy. When you get to 100+ fps, video almost starts to look more "real" than real life (partly because we're all so conditioned from a lifetime of 24, 50, and 60fps video, anything faster automatically triggers the "must be real" visual reflex).
And no, I'm not talking about "100hz", "120hz", "200hz", or "240hz" TVs. They aren't increasing the framerate -- they're just using the faster refresh rate with oversampling to simulate interlace blur and scanline fade (partly because high-quality motion-vector deinterlacing is hard to do well, and pretty much impossible to do in realtime, so the only alternative to 1080i60 video with bob and weave artifacts on inherently-progressive displays is to simulate the way CRT scanlines used to blur and fade.
And yes, I'm one of those weird people who likes to rip 1080i60 interlaced content, then let something like Avisynth with MVtools chew on it for a week to decompose each scene into a panning/zooming "background" with moving "sprites", then tween them and synthesize missing detail to make everything look like real 60fps progressive-scan live video. :-)
>It's all advertising.
Sorry, but no. No. A thousand times, "No!"
You can tell the difference between 480i and *anything* progressive on a 19" display across a smoky room the day after having laser eye surgery. It's the same difference you saw years ago when comparing any TV to any VGA monitor. It might have only been 640x480, but it had a pleasing "solid" appearance that was instantly visible compared to any TV.
That said, the difference between 480p60 and higher-res isn't quite as dramatic. 720p60 is a nice step up, but it's not night-and-day. 1080i60 can be better if there's not a lot of motion (to cause weave artifacts), but most of the time it gets bobbed into something with 960-1440 real horizontal pixels and 540 vertical pixels anyway. True 1080p60 is a sight to behold, but so little non-CGI real honest-to-god 1080p60 content exists, most of the time you're just seeing 1080p24 with every other frame shown 3 times in a row instead of 2, or maybe 1080i60 that's been deinterlaced with studio-grade postproduction hardware.
Also, "normal viewing distance" is so last-century. Real people with 60+ inch TVs sit almost close enough to touch the screen, because there's no *reason* to sit farther -- no screen door effect, no x-rays, and some nice perspective-filling immersion when you sit close to the screen.
> I was specifically commenting on your assertion that the market was demanding big cars,
> when in fact that market evaporated even before the economic crises.
Unless you count export markets for American cars. When somebody in Japan, Australia, Britain, Saudi Arabia, Russia, Mexico, or elsewhere in the world buys an imported American car, it's NOT going to be an American-made Ford Focus or Chevy Aveo... it's going to be a Mustang, a Corvette, an Escalade, or maybe even a big F-150 truck. *THOSE* are the American cars with viable export markets. If they were eliminated, America would basically cease to have cars anybody in a foreign country would be interested in buying. They have plenty of domestic econoboxes to choose from.
American automakers know what their niche market is, and it's a niche they understand well. Pre-bankruptcy, Chrysler developed a very cool hybrid turbo-electric drivetrain that used the electric traction motor with PID feedback to compensate for turbo lag. Slam the pedal to the floor, and the electric motor kicks in to add a few dozen horsepower while the turbine is spinning up. The net result is a sports car that gets ~20-24mpg instead of the usual 12-18mpg, and basically lets you have your cake & eat it too.
> the vast majority of SUV owners do not haul cargo, do not drive off-road, and do not frequently carry more than 3 passengers.
Apparently, you don't live in South Florida and drive home from work on freeways that have visible sheets of water, and neighborhood streets under INCHES of water after any meaningful rainstorm. This is our daily reality.
In Miami, a big SUV or pickup truck means the difference between being stranded for hours wherever you were when it started raining, and being able to hydroplane home at 70mph on the freeway in some degree of relative safety. As an individual, there's not much I can do to unclog the dysfunctional storm drains of Dade & Broward counties, and even less I can do to make it stop raining for days at a time. However, I can personally go out and buy a vehicle capable of driving through 8" of water without major risk to the engine block and mitigate 90% of the problems that would otherwise accompany months of endless rainstorms.
>But that would make Google...Evil.
Naw. More like Chaotic Neutral ;-)
> Am I missing something? What's the problem?
Big, huge bit of advice: don't link your Market developer account to a Gmail account that does anything besides sit passively and be an otherwise-unused Gmail account. If you associated your Market developer account with a gmail account associated with G+ and they suspended it, you'd be *beyond* screwed commercially. Depending upon how aggressively Google wanted to mess up, you might not even be allowed to *create* a new developer's account for Android Market (it would see your credit card number, do a credit check lookup to verify your identity, realize it was associated with a suspended developer's account, and most likely either refuse to allow its creation, or nuke it without refund within hours or days.
THAT's what people are screaming about -- the fact that Google is basically setting people up for domino-like failure cascades that could have devastating professional and commercial consequences, and there are more than a few scenarios where they could happen to somebody who's either outright innocent, or at least didn't deserve the equivalent of an online death penalty.
Well... if Google effectively revoked the purchase of everything I'd purchased through Android Market without refund or recourse by suspending my Gmail account (the only way to purchase anything through Android Market), I'd be pretty pissed and would probably sue them just as a matter of principle. It's one thing to take away somebody's email service. It's another thing to retroactively confiscate purchased software whose only available means of mainstream purchase is with that Gmail account (or, in the case of AT&T, the only way to purchase apps, period... at least for now).
Even if the apps remained working until the phone were reflashed or replaced, I reflash my phone frequently as is my inalienable Linus/RMS-given right. If Google wants to be able to kill Gmail accounts for any reason (or no reason at all), they'd better damn well come up with a parallel SSO identifier for Android phones that can survive Google Exile unscathed. ESPECIALLY given their "shoot first, then be disinterested in even pretending to care about asking questions later" policy. At this rate, by the end of the year, there will be at least a dozen movie plots circulating around Hollywood for a proposed movie called "GoogleX'ed" whose plot is somebody who gets exiled from Google and has his life cascade and snowball into disintegration around him as service after service gets cancelled because it depends upon some other service that ultimately was tied to a Google service.
Er, well, if you want to talk about "stereo" gear sold at K-mart... maybe. But if you stepped up even slightly (say, to Service Merchandise and Circuit City, late 80s) and considered Technics to be the low end baseline, things were much better.
Plus, let's not forget the coffee table sized monster subwoofers DAK used to sell and ship by motor freight. I still have one upstairs in the bonus room ;-)