Sorry, but Win2k was my favorite version of Windows, ever. It was the first and only version of Windows where you could literally do hot driver replacement of things like AGP miniport drivers or the entire video subsystem without having to reboot. Now, it seems like every new version of Windows tries to make you reboot if the installer so much as adds an item to Explorer's right-click context menu.
> Some of the most profitable smartphones in the industry have a non-removable battery and no microSD expansion slot.
Name one that isn't made by Apple.
Every Android phone I've ever seen that shipped with the fatal combo of "nonremovable battery and without microSD" basically sucked. The batteries were too small, and the lack of microSD just kicked it over the edge.
Here's a good reason to demand microSD: you CAN get your phone into a state where it's sufficiently damaged to make mounting it as a USB drive impossible, or into a bootloop where only a full factory wipe (including internal/sdcard) will get it out. When that happens, your data is gone. In contrast, it's amazingly hard to truly destroy a microSD card. Watch a disaster show on the Discovery Channel sometime, and observe how many firsthand accounts were filmed by dead videographers whose destroyed camcorder had a readable microSD card inside. The point is, even driving over your phone with a car is unlikely to render the data on the microSD card gone forever, unlike its regular internal flash (whose operation mostly depends upon having a phone that at least "kind of" still works).
My own phone? A Galaxy S3 with Seidio extended battery & active case. Total thickness is comparable to an Otterbox Defender, with approximately the protection of an Otterbox Metro. And I can say, it's damn nice to be able to use my phone all day without ever touching a battery charger or making the slightest attempt to conserve power. Give a phone a nice, big battery, replace the kernel governor with one that supports a variant of "Interactive" instead of "Ondemand", and your phone's lag problems just go away & it still runs all day on a single charge. And if that battery isn't quite enough (say, post-Hurricane), I have 3 more just like the one that came with the phone purchased from China for about 10 bucks as a cheap insurance policy. At that price, it's actually cheaper to use them once and throw them away than it would be to use a kludgy usb charger powered by alkaline batteries.
The same thing happens in Miami when a hotel or restaurant runs out of something, or a supplier falls through, and sends out employees to buy up every carton of vanilla & chocolate ice cream within 3-5 miles. Or 2-liter bottles of Coke (when their postmix beverage machine dies, and they run out to buy bottles of soft drinks to serve in the meantime). It doesn't happen nearly as often now thanks to Sam's Club, but it still happens every now and then. You'll go into Publix (the main grocery store chain here), and they'll have an aisle of something basic that no grocery store should ever be without that they'll just be totally wiped out of for no obvious reason.
Plenty of companies were founded by someone with the cash and means to quietly pull off some big purchase below the radar for pennies on the dollar to do something that would have been cost-prohibitive for a lumbering large corporation with purchasing department and 3 layers of bureaucracy. Like FedEx.
The whole thing that made FedEx possible was the semi-secret invention of a hush kit for the Boeing 727 by one of FedEx's founders. It enabled them to quietly buy up an entire fleet of relatively new jets for pennies on the dollar that were being liquidated at fire-sale prices by US airlines in anticipation of new noise restrictions. THEN, after they'd bought all the jets they wanted, they told everyone about the hushkit, and companies like UPS were out of luck because the 727 suddenly became viable for passenger use again for another 10-20 years and ceased to be dirt cheap.
It's easy, really... Google just has to find a way to smack down the incumbent network providers (AT&T, Comcast, etc) so they can't stop municipalities from laying their own fiber municipal networks with an access point every mile or so, then empower citizens to take matters directly into their own hands and lay their own fiber bundles in trenches they dig with their own shovels (or hang their own fiber from a public support attached to the lowest rung of the city's utility poles) to get to those access points if their HOA drags its feet or tries to tie them into proprietary services). Few things will put the fear of God into any HOA than being told their homeowners have the inalienable right to string wires from poles and dig their own trenches if they don't hurry up and get their own acts together. Get my fiber to a 10-gig switch with direct connectivity into NAP of the Americas in downtown Miami, and AT&T and Comcast can both go straight to hell as far as I'm concerned.
AFAIK, AT&T acquired Alltel spectrum, but Verizon acquired Alltel's customers, who were able to use Verizon's network without problems. There might have been a few oddball areas, but AFAIK, for Alltel customers, the change involved little more than new billing statements and a PRL update.
> how would this merger work? Tmobile is GSM whereas MetroPCS is CDMA.
The difference isn't as big as you think.
The difference between legacy TDMA-based 2G GSM/GPRS/EDGE and WCDMA-based 3G UMTS/HSPA+ is HUGE. Night and day. Literally, nothing in common besides a subset of the SIM card and the battery.
The difference between CDMA2000 voice/1xRTT data and EVDO is almost as big as the difference between 2G GSM and 3G UMTS/HSPA+.
The difference between CDMA2000 voice/1xRTT data and UMTS/HSPA+ is basically the presence of a SIM card, wider channels, and some evolutionary refinements. A tri-band (1700/1900/2100MHz) phone that does "everything" would cost about the same to make as a HSPA-only or CDMA2000-only phone that has to support 850 and 1900Mhz. Mode support is mainly a matter of IP licensing and firmware. What really drives up the cost is having to support 5 or more bands spread across the upper UHF and lower microwave spectrum.
Food for thought: the ONLY thing that prevents an AT&T iPhone 4S or Galaxy S2/S3 from working on Sprint or Verizon is radio firmware, FCC certification, and business policy. Google MDM6600 or MSM8960 sometime, or just read THIS for some gory details.
> Unless you're a DT or PCS shareholder, there's no reason to support this merger. None whatsoever.
Unless, of course, you're a MetroPCS customer with late-model Android phone (like the Galaxy S3) whose underlying chipset is perfectly capable of HSPA+ (assuming its soldered-in RUIM can be induced to act enough like a real SIM card to make a GSM network happy... I'm pretty sure they CAN, if push comes to shove...).
THOSE customers will absolutely be dancing in the streets, because it might mean they might get to start using T-Mobile's 6-14mbps HSPA+ network for data a few months from now, instead of limping along at ~2mbps or less on MetroPCS's EVDO, or roaming at a painful crawl on Sprint's third-world single-digit-kbps EVDO (slower than data anywhere on earth besides maybe rural Haiti). From what I've heard, even Metro's LTE is slower than T-Mobile's HSPA+.
If Metro's PCS voice spectrum is congested, and T-Mobile can show the FCC that they're actively using their own spectrum as well, I seriously doubt there's going to be any great push to force its divestiture unless some small regional carrier is looking *really* needy. I mean, who's the FCC going to make them sell it to? AT&T or Verizon? Please. Sprint doesn't really need more 1900MHz spectrum, doesn't want AWS spectrum, and couldn't afford to buy it anyway.
Remember, most of the merger-forced spectrum sales over the past 10 years have been *TO* T-Mobile, because they're the ones who've always been starving for 1900MHz spectrum.
No and yes. They'll "Go Canadian" and do what carriers in Canada did -- keep using circuit-switched CDMA for voice (since it's built out, a sunk cost, and already "there"), and new phones will default to HSPA+ & LTE for data (falling back to EVDO, 1xRTT, EDGE, and GPRS when HSPA+/LTE doesn't exist).
In ~2 years, they can start scaling back EVDO and reallocate spectrum to LTE and HSPA+. Another year or two past that, and they can eliminate EVDO entirely (keeping 1xRTT for the few stragglers who won't notice the difference anyway, since it can coexist with circuit-switched CDMA voice). They can consolidate cell sites and share trunk lines between CDMA & GSM voice, and gradually scale back CDMA voice/1xRTT to the minimum required for a single channel.
Put another way, the end of Metro CDMA will be like the end of the Roman Empire... there won't be a grand "flip the switch and shut it all down" day, so much as fading into irrelevance before parts start to just break and stay broken because nobody cares enough to fix them anymore. The amount of money it would cost to openly and officially shut down CDMA is more than it would cost to just make it moot going forward and progressively neglect it into oblivion, because the users who'll hang on to their old phones for 10 years won't notice the difference between EVDO and 1xRTT anyway.
In the long run, recipient pays was actually GOOD for consumers.
In countries where callers had to pay, mobile networks had no real incentive to lower the cost of incoming calls, other than maybe allowing free calls between their own customers. In fact, they had every incentive to shift the bulk of their charges to incoming callers, because they don't have the option of taking their business elsewhere if they think the charges are excessive. Their only option is to not call you at all, which is rarely a viable option.
In contrast, Americans were kind of screwed for incoming call charges for most of the 90s, but once the costs started coming down, carriers found themselves under immense competitive pressure to make the first incoming minute free, increase the number of included minutes, make nights/weekends free (or at least offer more minutes than 98% of customers could ever possibly use), then make nights begin earlier and end later. With "recipient pays", the customer is the one paying both ends of the call, so they're going to scream about ALL of the charges... not just the ones they personally have to pay.
Given the culture and habits of American carriers, let's look at what life would likely be like today if America had been "caller pays" all along:
* If you're a Sprint customer, calling other Sprint customers is free. If you're an AT&T customer, calling other AT&T customers is free. Ditto, and exactly the same, for T-Mobile, Verizon, US Cellular, Metro PCS, and the rest.
* If you're a Sprint customer, calls to an AT&T landline customer are 21c/minute. Calls to an AT&T wireless customer are 39c/minute. Calls to a Verizon customer are 37c/minute in the northeast, and 44c/minute in Nevada, Idaho, and the Seattle metro area. Calls to a T-Mobile customer are 13c/minute, unless the mobile user is roaming on AT&T's network, in which case the call will cost 52c/minute (13c to T-Mobile, 39c to AT&T).
* If you're an AT&T customer, calls to Verizon customers are 5c/minute, and vice-versa. But calls to Sprint are 21c. Calls to T-Mobile customers are temporarily free for the next 19 months as a post-merger condition, but will thereafter return to 24c/minute.
* Verizon customers pay more or less the same charges as AT&T customers, except for some unknown reason, Verizon customers get charged a whopping 88c/minute when calling MetroPCS customers.
At the end of the day, customers who use the same company as everyone they call with any regularity might pay slightly lower rates, but most of us would just get totally screwed by everyone else's carriers. The most popular apps in Android Market would be alternate dialers that dynamically look up the current cost of calling a given number before you press 'send', and warn you of the exact cost based upon the recipient's current carrier... or place the call through one of 5 alternate VoIP providers, if they end up being slightly cheaper per minute to call a given number than calling them directly. In other words, we'd be screwed at least as badly overall as we are now, but we'd be getting screwed by more companies in more complicated ways.
Actually, we're both kind of right. I was thinking more specifically of programmed I/O, which would absolutely outstrip even the fastest Intel-architecture CPU at 10gbps speeds, and I forgot that you wouldn't have to actually touch every single byte with the CPU... in real life, you'd have a DMA controller to buffer bits from the wire while the CPU slogged along and parsed the first few bytes, then the CPU would tell another DMA controller how to dispatch the bytes in the buffer that continued to accumulate in the background while the CPU parsed the header. Nevertheless, I think it's an important reality check to point out when people start to get jaded & think 10gbps is "slow".
This is the same reason why lots of consumer-grade routers crash when users try to torrent over DOCSIS3.0 cable internet. They just plain can't touch every byte and do realtime NAT at the data speeds the newest cable modems are capable of sustaining (at least, until Comcast starts throttling you). Eventually, they run out of ram, or they get so backlogged that the remote server starts to resend data that's actually sitting in the incoming buffer and hasn't even been looked at yet by the router. They can deal with brief bursts of data from web pages, but an unending surge of megabyte after megabyte eventually pushes them over the edge unless the remote server sees what's happening, then backs off and throttles itself.
At the rate we're going, "8P8C" in the terabit+ category will probably end up meaning "cable with 4 pairs of single-mode fibers". When you start talking about terahertz signaling rates, a single fiber starts looking like a pair of copper wires & you start to feel like if it hasn't quite outstripped the final viable limits of what it can do, it's getting pretty damn close.
As a practical matter, wire speeds faster than 10gbps almost *have* to be treated like parallel bundles of fast, but independent bitstreams that happen to be sharing a common transmission medium, but are all traveling in parallel & are oblivious to each other's content, just because you eventually have to switch or route the traffic, and there's a practical limit to the speeds even the fastest DSP-like purpose-built CPU can achieve on-die, let alone on its circuit board.
Aggregating 10 1-gigabit quasi-serial links into 10 independent streams that might be modulated together into a single fiber, then later demodulated and restored to 10 1-gigabit quasi-serial links is one thing. Trying to actually switch (let alone route) a true single-10gbps bitstream by studying its IP header and making decisions based upon things like its ipv6 address is another matter entirely.
I met somebody about a year or two ago who told me that routing 10-gigabit traffic today is kind of like sexing newborn chickens. At that speed, you aren't analyzing headers... you're making single-bit snap judgments on a slightly-blurry bitstream, and hoping it wasn't noise that sent a datagram meant for someone in Ohio to Shanghai instead. Or more precisely, you have a few circuits sniffing the blur in parallel, voting on what they think its destination is likely to be, and majority rule deciding where it goes next.
Putting into perspective just how fast 10gbps is from the perspective of a single user, in the time it takes the fastest Intel-architecture AMD64 CPU money can buy today to test a single byte already in a register and determine whether its value is zero or nonzero, an entire byte or more would fly by on the 10gbps wire.
> If it's so profitable to build a telecommunications company then why are more local ones not popping up and serving our desires?
The fact that AT&T and Comcast will have every lobbyist and lawyer on their payroll swarming the city/county commission, state regulators, and anyone else they can think of to get the local government authority prohibit you from doing it? Kind of like they do every time some uppity community gets fed up and decides to lay its own fiber?
You know, it's scary. If Union Pacific, Western Union, Alexander Graham Bell, and Westinghouse had the accounting values and practices of modern American corporations, we'd still be a third-world country with the infrastructure of Zaire.
Can you imagine what we would have ended up with in Orlando if Walt Disney had been beholden to Wall Street's insatiable demands for immediate payoff? I mean, my god... Reedy Creek Improvement District was so rural when Walt's shell companies closed on the final deed, every square inch of the Magic Kingdom -- including its "wild" parts -- was scraped bare down to the limestone and literally terraformed into what exists today. They even built their own MONORAIL. Wow. Try getting THAT past a group of passionless institutional investors whose only question concerns its expected ROI. If modern Wall Street had its way, we'd be parking in Goofy 47, then riding a bus to the entrance instead of a monorail. After all, it would reduce costs by 2.7 cents per guest per day!
Well... except for the fact that it would have made the Magic Kingdom a lot less "magical" to kids like us, and we would have thrown a fit and begged to go to Bush Gardens instead (yeah, I'll admit it... my parents could have easily saved a few hundred bucks each summer by skipping the Magic Kingdom, getting a room at the Contemporary Resort, and letting me ride the monorail all day with complete legitimacy as a hotel guest instead. But I digress...)
Even if you assume 100% of that $4k was capital construction cost, it's something that's going to have a useful life measured in centuries. Plus, much of that cost was self-inflicted by Verizon itself. If Verizon were smart, they would have pulled new fiber everywhere, left the copper in place, then spun off a new company to own the copper wires (but not the easements through which they ran) and sell landline phone service & discount DSL to poor people. The new company would have never been a long-term threat to Verizon (with no ROW of their own, and only an easement to run copper UTP, they'd have been hitting their head against the ceiling within a decade or two), and the value to a buyer would have been far more than the scrap value of the copper wires and lost customers that Verizon wants to ditch anyway.
Almost, but not quite. There are plenty of Android phones with bootloaders that are unlocked (officially or otherwise), but are still stuck with old kernels because they depend upon binary loadable kernel modules that are not themselves open-source. Remember, Linux doesn't have a stable ABI, so loadable kernel modules ("drivers", in Windows parlance) are specific to a kernel version.
This is probably the #1 source of recurring grief at xda-developers.com. Every new version of Android ships with a new kernel that breaks every binary driver (camera, GPS, wi-fi, sensors, bluetooth, etc) that came before it. IMHO, it's the #1 reason why Google needs to just fork Android's kernel (since 99.9% of the changes are Android-specific, and few mainline changes really matter to Android devices) and commit to having either a stable ABI, or at least come up with a reasonable thunking layer to enable binary modules that are a year or two old to continue working with newer kernels, absent some *really* good reason (like a catastrophically bad security vulnerability) to justify a compatibility-breaking change.
Some people might not like voiding their warranty the day they buy their phone.
Which is why we all make a nandroid backup before flashing a new firmware.
> Some people might not like voiding their warranty the day they buy their phone.
Manufacturers can lie about warranty-invalidation until they're blue in the face. The Magnuson-Moss Warranty Act ( http://en.wikipedia.org/wiki/Magnuson%E2%80%93Moss_Warranty_Act ) is a potent weapon that no manufacturer, not even Apple in their most arrogant AT&T-exclusive hissy fit, would dare to push back against because the FTC will smack them down and make a total example out of them.
Under Magnuson-Moss, a manufacturer can only deny warranty coverage if they can demonstrate that whatever the consumer did was literally the cause of the failure... and historically, the FTC hasn't made their job easy. They basically get one chance to make their case to the FTC, and if the FTC thinks the company is harassing the customer and wasting their time on a silly excuse, it will instantly smack them down and hit them with a huge fine.
If the manufacturer wanted to use "we had to reflash it via JTAG to stock" as an excuse for denying the claim or imposing a service fee, they'd have to testify that they don't routinely JTAG-reflash to stock as a troubleshooting step anyway.
If they tried to argue that you somehow triggered a condition via software that caused damage (say, setting a pair of directly-connected GPIO pins to outputs, with one high and one low), they'd still be backed up against the wall and told they were idiots for not putting a resistor between them, or at least going out of their way to make it abundantly clear to end users that custom firmware must never, ever do that specific action. In stark contrast to most consumer non-law, the FTC takes consumer rights seriously, and doesn't take crap from companies who try to wave vague disclaimers around and use them as an excuse and blanket license to run roughshod over consumers. The barrier isn't quite insurmountable, but a company that tried to fight it would have an uphill battle, and quickly discover that its usual dirty tricks weren't going to work this time around.
Companies doing dirty tricks with warranty coverage is nothing new. The same things phone manufacturers try to do today, American automakers did to our parents and grandparents openly and proudly, with a dash of extra salt to rub into consumer wounds ~30 years ago.
Magnuson-Moss is a rare gem of consumer-protection law passed by an angry congress fed up with the increasingly-bold abuses of the 3 most powerful companies in America at the time. Apple, Samsung, HTC, and Motorola might be powerful... but they're *nothing* compared to the "Big Three" American automakers circa 1975, and they know it.
Unfortunately, it's NOT against the law for a company to blatantly lie about its legal responsibilities, so companies can say anything and put all the restrictive text they want to put in their warranty descriptions. You just have to know that when push comes to shove, all you have to do is whisper the magic phrase "Magnuson-Moss" to get your complaint *instantly* escalated to the most senior manager on site and get total white-glove treatment and profuse apologies for the "misunderstanding" (inevitably blamed on the tier-1 support staff, who were just doing what the script told them to do).
Stackoverflow is by far the most tolerant of questions that don't fit neatly into a perfect box that can be wrapped and declared to be answered. Questions that deviate from the formula DO eventually get moderated in a way that allows them to fade from view and get pushed to the bottom of search results, but there's (fortunately) very little of the religious zeal for stamping out Badthink that seems to infect most of its other sites.
In contrast to SO, the other SE sites really seem to attract moderators who fall into the "ISTJ" personality type (strong sense of right & wrong, disdain for hypothetical problems or theoretical solutions compared to concrete solutions to clearly-defined problems, deference to authority and tradition, sense that the destination or goal is the whole reason for doing something, and almost zero tolerance for "inefficiency"). As luck would have it, this group tends to clash rather badly with the INTP (and weakly-J INTJ) personality types that are common among programmers, and StackOverflow is the Jerusalem where the two share an uneasy & tense coexistence... surrounded by a StackExchange ocean of more public and visible conflict(*).
By comparison, ServerFault is *hardcore* ISTJ territory, and the INTx-ISTJ conflict is extremely visible. Many SF mods act like it's their religious duty to stamp out even the slightest hint of something that might be interpreted as circumvention of official authority (or even just frowned upon). To an ISTJ, rules are meant to be followed and enforced... including blanket rules with awkward edge cases, if carving out exceptions for those edge cases would turn a rule that can be concisely expressed by a single sentence into 27 pages of flowcharts, BNF grammars, and appendices. In contrast, the INTx (particularly INTPs) will meticulously document the scenarios where the "one size fits nobody" rule produces absurd results, then go nuclear when the ISTJs say they don't care and it doesn't matter.
Ultimately, what SE in general needs is a Wikipedia-like "discussion" mechanism to move the discussions a step away from the factual essence of the main page, but nevertheless recognize that healthy disagreement and discussion can be valuable, especially for topics where yesterday's best-practice might be today's horrific security vulnerability or architectural dead end. The fact is, there are lots of problems that really *don't* have viable solutions (yet), and documentation of the specific roadblocks can be valuable in itself.
(*)For anyone who cares, the MBTI personality type most likely to clash with yours is NOT necessarily its polar opposite. The truly epic FIGHTS tend to occur between two individuals of the SAME introversion/extroversion type, but conflict with respect to the remaining 3 attributes. An INTP and INTJ with common interests might be new best friends for the first 10 minutes of a 4 hour car trip, and bitter enemies by the end.
That's a mostly accurate summary of the outwardly visible effects a nontechnical end user who only wants to "surf and talk" might see, but it doesn't quite describe the actual problem. If you know exactly what's going on behind the scenes, there ARE ways to do simultaneous voice + data on Sprint (though most of them require spending more money for thirdparty services and doing an end-run around Sprint itself).
Let's start with voice calls. If you want to terminate a voice call through Sprint, there's exactly one way it can happen: via circuit-switched CDMA. The call travels between Sprint and whomever is at the other end of the call over the public switched telephone network (PSTN). Between Sprint's switching center and the phone, the call is transported via CDMA2000 voice.
No Sprint phone I'm aware of can simultaneously handle an active CDMA voice call and IP data using either EVDO or 1xRTT. It's physically impossible due to the way both are implemented on Sprint's phones. HOWEVER, in most of Sprint's recent phones, there's no hardware reason why you can't have an active voice call AND use data via Wimax, LTE, and/or wifi. Some Sprint phones have shipped with firmware that disabled data use during voice calls, but that was mainly due to tech support and demographics. Basically, they didn't feel like dealing with less-technical users who couldn't be assumed to understand the difference between the different data modes, be aware of their connectivity from minute-to-minute, and actively manage their data connection mode in order to do both at once.
That said, there's nothing (besides battery life and added subscription cost) to stop you from acquiring VoIP service from someone besides Sprint, running a SIP client on your phone, and making simultaneous VoIP calls while using any working data mode, including EVDO and (maybe, bandwidth permitting) 1xRTT. You can even use call forwarding to forward your incoming Sprint calls to the VoIP number. The downside is that your phone will drain the battery a LOT faster, because you'll have to actively poll for incoming calls (normal incoming calls are handled by having the phone poll Sprint's towers via the same mechanism used for text messages; with VoIP, the phone is polling twice as much, and has to maintain an active data connection to do it). One compromise is to make your outgoing calls via VoIP, but let your incoming calls continue to come in through Sprint (losing EVDO/1xRTT data connectivity when it happens).
So... you might be wondering... if end users can get simultaneous voice+data, even over CDMA data modes, by using VoIP service acquired independently of Sprint... why can't Sprint itself do it? Basically, their switching equipment can't handle it. Nothing that couldn't mostly be hacked around, but it would have been expensive, likely to cause problems (witness the thrashing many wimax Sprint phones do when they can't make up their mind between EVDO and wimax, and just keep breaking the data connection and thrashing wildly between the two), and would have still left Sprint with compromises compared to HSPA+.
For what it's worth, these problems are nothing new... GSM networks went through the exact same hardware problem 10 years ago when they transitioned from TDMA-based GSM/GPRS/EDGE to WCDMA-based UMTS/HS(D|U)PA(+). The main difference is that European phone companies actually WERE able to buy off the shelf switching equipment to deal with it, whereas Sprint would have had to cobble its own half-baked solution in-house. Qualcomm WAS actually working on SVDO to replace EVDO (doing more or less the same thing as HSPA+), but most CDMA carriers in other countries decided to skip EVDO and just transition to UMTS/HS(D|U)PA(+) instead. Sprint and Verizon couldn't do that, because they didn't have the pair of 10MHz uplink and downlink channels in all of their markets.
Sprint AND Verizon together would have been enough to motivate Qualcomm to finish development of SVDO, but Verizon hoped to strike a deathblow against Sprint by formally aband
Actually, it's more like the government forced lenders to drop their standards for certain favored groups and backed it up with federal insurance, then the lenders realized they could just drop their standards for everyone to the same level and wash their hands of its consequences. The same loopholes that allowed a struggling working-poor person to qualify for a home with zero down and an artfully-contrived assessment of his or her assets & annual income allowed middle-class wannabe-investors to buy pre-construction condos with zero down for double their realistic value in the hopes that they'd be able to flip them for 4 times their realistic value. This inflated prices in general, and got otherwise-sane middle-class buyers buying a home for themselves to agree to stretch their finances to the breaking point to afford to buy a house that everyone thought would become unaffordable to them forever if they didn't buy immediately. In cities like Miami, houses that would have barely gotten $80,000 in 1995 were bought sight-unseen for $250,000 in 2003, and $399,000 in 2005, by speculators planning to demolish them and build a million-dollar replacement to flip in their place. Caught in the middle was the example middle-class family facing the prospect being priced out of the housing market forever.
The only, and I mean *ONLY* thing that kept the market from going completely nuclear, were the guidelines that made it hard to qualify for a loan of more than ~$360k with zero down, and almost impossible to qualify for a loan of more than ~$460k without real proof that you could afford it and a major down payment. For a time, both were de-facto price caps in South Florida... tiny 650sf condos with no reserved parking were almost universally $360,000, two-bedroom townhomes (or single-family homes in borderline-ghetto neighborhoods) were almost universally $460k, and everything else involved $360k or $460k plus $50k-100k cash down. Without those limits, there would have absolutely been people buying zero-down million-dollar 1 bedroom condos, despite their annual incomes being less than just a year's worth of INTEREST payments on the loan.
> I am yet to see DNS fail badly. I have seen plenty of people who don't understand it say it does, when the problem is invariably routing or a firewall.
Note the key phrase, "who don't understand it" and its modifiers "routing or a firewall". There's a HELL of a lot of people who happen to fall into that category, and whose frustration goes off the scale when something fails to work because the slightest configuration problem will break it, and if you manage to avoid a subtle semantic bug in a zonefile somewhere, factors upstream that are beyond your control can still break it in ways that are almost impossible to distinguish from that same hypothetical zonefile bug. Bind is a cruel, heartless, sadistic, and demanding master. I struggle to think of anything that universally strikes fear in the heart of otherwise brave men than "Can you set up the DNS for us from scratch? We registered the domain yesterday. The computer's over there, and here's the Ubuntu installation DVD."
DNS only "just works" when some OTHER unfortunate soul has already set it up and spent the day troubleshooting it for you. Worst of all, there's a big, fuzzy gray area of "works for me, and apparently for him, but not for you for some unknown reason".
So, yes... for 17 or 18 artisan-level gurus who've achieved Englightenment, DNS is easy and straightforward to set up. For the other 99.9% of individuals unfortunate enough to find themselves tasked with the duty of setting it up at the server end, it's pure hell, and Bind was a punishment invented by God for sadistic entertainment purposes.
When I talk about "DNS Failure", I'm not talking about it from the perspective of pure end users who connect to a network and make use of DNS that somebody else has already gotten to work for them. I'm talking about the unrelenting hell of being someone who lacks control of his upstream network configuration trying to set up his own DNS server and make sense of errors that could be caused by just about anything, or (almost) nothing at all, and can take a relative eternity to troubleshoot when it happens.
You're right, but what's the pixel resolution of a "720p class" TV, and how big is its "19 inch class" screen?
The situation I'm complaining about is what it would be like if that 2x4 piece of wood were a "2x4 class board product" whose actual dimensions could be anywhere between 1 x 2.5 to 1.4 x 3.5, and endless racks of 70-cent "board products" at Home Depot made a real, honest to god 2x4 cost $15 and have to be special-ordered.
The problem is, DNS is like USB, and IPv4 is like RS232. If you're anywhere close to being right, you can probably get ipv4 (or a real serial port set to 9600-8-N-1) to work well enough to give you clues about what the real problem is. In contrast, DNS (like USB) tends to just fail hard and catastrophically, giving no obvious clues about what might actually be wrong.
Slashdot Mirror
Sorry, but Win2k was my favorite version of Windows, ever. It was the first and only version of Windows where you could literally do hot driver replacement of things like AGP miniport drivers or the entire video subsystem without having to reboot. Now, it seems like every new version of Windows tries to make you reboot if the installer so much as adds an item to Explorer's right-click context menu.
> Some of the most profitable smartphones in the industry have a non-removable battery and no microSD expansion slot.
Name one that isn't made by Apple.
Every Android phone I've ever seen that shipped with the fatal combo of "nonremovable battery and without microSD" basically sucked. The batteries were too small, and the lack of microSD just kicked it over the edge.
Here's a good reason to demand microSD: you CAN get your phone into a state where it's sufficiently damaged to make mounting it as a USB drive impossible, or into a bootloop where only a full factory wipe (including internal /sdcard) will get it out. When that happens, your data is gone. In contrast, it's amazingly hard to truly destroy a microSD card. Watch a disaster show on the Discovery Channel sometime, and observe how many firsthand accounts were filmed by dead videographers whose destroyed camcorder had a readable microSD card inside. The point is, even driving over your phone with a car is unlikely to render the data on the microSD card gone forever, unlike its regular internal flash (whose operation mostly depends upon having a phone that at least "kind of" still works).
My own phone? A Galaxy S3 with Seidio extended battery & active case. Total thickness is comparable to an Otterbox Defender, with approximately the protection of an Otterbox Metro. And I can say, it's damn nice to be able to use my phone all day without ever touching a battery charger or making the slightest attempt to conserve power. Give a phone a nice, big battery, replace the kernel governor with one that supports a variant of "Interactive" instead of "Ondemand", and your phone's lag problems just go away & it still runs all day on a single charge. And if that battery isn't quite enough (say, post-Hurricane), I have 3 more just like the one that came with the phone purchased from China for about 10 bucks as a cheap insurance policy. At that price, it's actually cheaper to use them once and throw them away than it would be to use a kludgy usb charger powered by alkaline batteries.
The same thing happens in Miami when a hotel or restaurant runs out of something, or a supplier falls through, and sends out employees to buy up every carton of vanilla & chocolate ice cream within 3-5 miles. Or 2-liter bottles of Coke (when their postmix beverage machine dies, and they run out to buy bottles of soft drinks to serve in the meantime). It doesn't happen nearly as often now thanks to Sam's Club, but it still happens every now and then. You'll go into Publix (the main grocery store chain here), and they'll have an aisle of something basic that no grocery store should ever be without that they'll just be totally wiped out of for no obvious reason.
Plenty of companies were founded by someone with the cash and means to quietly pull off some big purchase below the radar for pennies on the dollar to do something that would have been cost-prohibitive for a lumbering large corporation with purchasing department and 3 layers of bureaucracy. Like FedEx.
The whole thing that made FedEx possible was the semi-secret invention of a hush kit for the Boeing 727 by one of FedEx's founders. It enabled them to quietly buy up an entire fleet of relatively new jets for pennies on the dollar that were being liquidated at fire-sale prices by US airlines in anticipation of new noise restrictions. THEN, after they'd bought all the jets they wanted, they told everyone about the hushkit, and companies like UPS were out of luck because the 727 suddenly became viable for passenger use again for another 10-20 years and ceased to be dirt cheap.
It's easy, really... Google just has to find a way to smack down the incumbent network providers (AT&T, Comcast, etc) so they can't stop municipalities from laying their own fiber municipal networks with an access point every mile or so, then empower citizens to take matters directly into their own hands and lay their own fiber bundles in trenches they dig with their own shovels (or hang their own fiber from a public support attached to the lowest rung of the city's utility poles) to get to those access points if their HOA drags its feet or tries to tie them into proprietary services). Few things will put the fear of God into any HOA than being told their homeowners have the inalienable right to string wires from poles and dig their own trenches if they don't hurry up and get their own acts together. Get my fiber to a 10-gig switch with direct connectivity into NAP of the Americas in downtown Miami, and AT&T and Comcast can both go straight to hell as far as I'm concerned.
AFAIK, AT&T acquired Alltel spectrum, but Verizon acquired Alltel's customers, who were able to use Verizon's network without problems. There might have been a few oddball areas, but AFAIK, for Alltel customers, the change involved little more than new billing statements and a PRL update.
> how would this merger work? Tmobile is GSM whereas MetroPCS is CDMA.
The difference isn't as big as you think.
The difference between legacy TDMA-based 2G GSM/GPRS/EDGE and WCDMA-based 3G UMTS/HSPA+ is HUGE. Night and day. Literally, nothing in common besides a subset of the SIM card and the battery.
The difference between CDMA2000 voice/1xRTT data and EVDO is almost as big as the difference between 2G GSM and 3G UMTS/HSPA+.
The difference between CDMA2000 voice/1xRTT data and UMTS/HSPA+ is basically the presence of a SIM card, wider channels, and some evolutionary refinements. A tri-band (1700/1900/2100MHz) phone that does "everything" would cost about the same to make as a HSPA-only or CDMA2000-only phone that has to support 850 and 1900Mhz. Mode support is mainly a matter of IP licensing and firmware. What really drives up the cost is having to support 5 or more bands spread across the upper UHF and lower microwave spectrum.
Food for thought: the ONLY thing that prevents an AT&T iPhone 4S or Galaxy S2/S3 from working on Sprint or Verizon is radio firmware, FCC certification, and business policy. Google MDM6600 or MSM8960 sometime, or just read THIS for some gory details.
> Unless you're a DT or PCS shareholder, there's no reason to support this merger. None whatsoever.
Unless, of course, you're a MetroPCS customer with late-model Android phone (like the Galaxy S3) whose underlying chipset is perfectly capable of HSPA+ (assuming its soldered-in RUIM can be induced to act enough like a real SIM card to make a GSM network happy... I'm pretty sure they CAN, if push comes to shove...).
THOSE customers will absolutely be dancing in the streets, because it might mean they might get to start using T-Mobile's 6-14mbps HSPA+ network for data a few months from now, instead of limping along at ~2mbps or less on MetroPCS's EVDO, or roaming at a painful crawl on Sprint's third-world single-digit-kbps EVDO (slower than data anywhere on earth besides maybe rural Haiti). From what I've heard, even Metro's LTE is slower than T-Mobile's HSPA+.
If Metro's PCS voice spectrum is congested, and T-Mobile can show the FCC that they're actively using their own spectrum as well, I seriously doubt there's going to be any great push to force its divestiture unless some small regional carrier is looking *really* needy. I mean, who's the FCC going to make them sell it to? AT&T or Verizon? Please. Sprint doesn't really need more 1900MHz spectrum, doesn't want AWS spectrum, and couldn't afford to buy it anyway.
Remember, most of the merger-forced spectrum sales over the past 10 years have been *TO* T-Mobile, because they're the ones who've always been starving for 1900MHz spectrum.
> My only question is, will Metro go GSM?
No and yes. They'll "Go Canadian" and do what carriers in Canada did -- keep using circuit-switched CDMA for voice (since it's built out, a sunk cost, and already "there"), and new phones will default to HSPA+ & LTE for data (falling back to EVDO, 1xRTT, EDGE, and GPRS when HSPA+/LTE doesn't exist).
In ~2 years, they can start scaling back EVDO and reallocate spectrum to LTE and HSPA+. Another year or two past that, and they can eliminate EVDO entirely (keeping 1xRTT for the few stragglers who won't notice the difference anyway, since it can coexist with circuit-switched CDMA voice). They can consolidate cell sites and share trunk lines between CDMA & GSM voice, and gradually scale back CDMA voice/1xRTT to the minimum required for a single channel.
Put another way, the end of Metro CDMA will be like the end of the Roman Empire... there won't be a grand "flip the switch and shut it all down" day, so much as fading into irrelevance before parts start to just break and stay broken because nobody cares enough to fix them anymore. The amount of money it would cost to openly and officially shut down CDMA is more than it would cost to just make it moot going forward and progressively neglect it into oblivion, because the users who'll hang on to their old phones for 10 years won't notice the difference between EVDO and 1xRTT anyway.
> Your calling me. Why should I have to pay?
In the long run, recipient pays was actually GOOD for consumers.
In countries where callers had to pay, mobile networks had no real incentive to lower the cost of incoming calls, other than maybe allowing free calls between their own customers. In fact, they had every incentive to shift the bulk of their charges to incoming callers, because they don't have the option of taking their business elsewhere if they think the charges are excessive. Their only option is to not call you at all, which is rarely a viable option.
In contrast, Americans were kind of screwed for incoming call charges for most of the 90s, but once the costs started coming down, carriers found themselves under immense competitive pressure to make the first incoming minute free, increase the number of included minutes, make nights/weekends free (or at least offer more minutes than 98% of customers could ever possibly use), then make nights begin earlier and end later. With "recipient pays", the customer is the one paying both ends of the call, so they're going to scream about ALL of the charges... not just the ones they personally have to pay.
Given the culture and habits of American carriers, let's look at what life would likely be like today if America had been "caller pays" all along:
* If you're a Sprint customer, calling other Sprint customers is free. If you're an AT&T customer, calling other AT&T customers is free. Ditto, and exactly the same, for T-Mobile, Verizon, US Cellular, Metro PCS, and the rest.
* If you're a Sprint customer, calls to an AT&T landline customer are 21c/minute. Calls to an AT&T wireless customer are 39c/minute. Calls to a Verizon customer are 37c/minute in the northeast, and 44c/minute in Nevada, Idaho, and the Seattle metro area. Calls to a T-Mobile customer are 13c/minute, unless the mobile user is roaming on AT&T's network, in which case the call will cost 52c/minute (13c to T-Mobile, 39c to AT&T).
* If you're an AT&T customer, calls to Verizon customers are 5c/minute, and vice-versa. But calls to Sprint are 21c. Calls to T-Mobile customers are temporarily free for the next 19 months as a post-merger condition, but will thereafter return to 24c/minute.
* Verizon customers pay more or less the same charges as AT&T customers, except for some unknown reason, Verizon customers get charged a whopping 88c/minute when calling MetroPCS customers.
At the end of the day, customers who use the same company as everyone they call with any regularity might pay slightly lower rates, but most of us would just get totally screwed by everyone else's carriers. The most popular apps in Android Market would be alternate dialers that dynamically look up the current cost of calling a given number before you press 'send', and warn you of the exact cost based upon the recipient's current carrier... or place the call through one of 5 alternate VoIP providers, if they end up being slightly cheaper per minute to call a given number than calling them directly. In other words, we'd be screwed at least as badly overall as we are now, but we'd be getting screwed by more companies in more complicated ways.
Actually, we're both kind of right. I was thinking more specifically of programmed I/O, which would absolutely outstrip even the fastest Intel-architecture CPU at 10gbps speeds, and I forgot that you wouldn't have to actually touch every single byte with the CPU... in real life, you'd have a DMA controller to buffer bits from the wire while the CPU slogged along and parsed the first few bytes, then the CPU would tell another DMA controller how to dispatch the bytes in the buffer that continued to accumulate in the background while the CPU parsed the header. Nevertheless, I think it's an important reality check to point out when people start to get jaded & think 10gbps is "slow".
This is the same reason why lots of consumer-grade routers crash when users try to torrent over DOCSIS3.0 cable internet. They just plain can't touch every byte and do realtime NAT at the data speeds the newest cable modems are capable of sustaining (at least, until Comcast starts throttling you). Eventually, they run out of ram, or they get so backlogged that the remote server starts to resend data that's actually sitting in the incoming buffer and hasn't even been looked at yet by the router. They can deal with brief bursts of data from web pages, but an unending surge of megabyte after megabyte eventually pushes them over the edge unless the remote server sees what's happening, then backs off and throttles itself.
At the rate we're going, "8P8C" in the terabit+ category will probably end up meaning "cable with 4 pairs of single-mode fibers". When you start talking about terahertz signaling rates, a single fiber starts looking like a pair of copper wires & you start to feel like if it hasn't quite outstripped the final viable limits of what it can do, it's getting pretty damn close.
As a practical matter, wire speeds faster than 10gbps almost *have* to be treated like parallel bundles of fast, but independent bitstreams that happen to be sharing a common transmission medium, but are all traveling in parallel & are oblivious to each other's content, just because you eventually have to switch or route the traffic, and there's a practical limit to the speeds even the fastest DSP-like purpose-built CPU can achieve on-die, let alone on its circuit board.
Aggregating 10 1-gigabit quasi-serial links into 10 independent streams that might be modulated together into a single fiber, then later demodulated and restored to 10 1-gigabit quasi-serial links is one thing. Trying to actually switch (let alone route) a true single-10gbps bitstream by studying its IP header and making decisions based upon things like its ipv6 address is another matter entirely.
I met somebody about a year or two ago who told me that routing 10-gigabit traffic today is kind of like sexing newborn chickens. At that speed, you aren't analyzing headers... you're making single-bit snap judgments on a slightly-blurry bitstream, and hoping it wasn't noise that sent a datagram meant for someone in Ohio to Shanghai instead. Or more precisely, you have a few circuits sniffing the blur in parallel, voting on what they think its destination is likely to be, and majority rule deciding where it goes next.
Putting into perspective just how fast 10gbps is from the perspective of a single user, in the time it takes the fastest Intel-architecture AMD64 CPU money can buy today to test a single byte already in a register and determine whether its value is zero or nonzero, an entire byte or more would fly by on the 10gbps wire.
> If it's so profitable to build a telecommunications company then why are more local ones not popping up and serving our desires?
The fact that AT&T and Comcast will have every lobbyist and lawyer on their payroll swarming the city/county commission, state regulators, and anyone else they can think of to get the local government authority prohibit you from doing it? Kind of like they do every time some uppity community gets fed up and decides to lay its own fiber?
> hardly anyone really needs high speed Internet at home. Many just want it.
"The telephone is a wonderful invention. I foresee the day when every major city will have one!"
You know, it's scary. If Union Pacific, Western Union, Alexander Graham Bell, and Westinghouse had the accounting values and practices of modern American corporations, we'd still be a third-world country with the infrastructure of Zaire.
Can you imagine what we would have ended up with in Orlando if Walt Disney had been beholden to Wall Street's insatiable demands for immediate payoff? I mean, my god... Reedy Creek Improvement District was so rural when Walt's shell companies closed on the final deed, every square inch of the Magic Kingdom -- including its "wild" parts -- was scraped bare down to the limestone and literally terraformed into what exists today. They even built their own MONORAIL. Wow. Try getting THAT past a group of passionless institutional investors whose only question concerns its expected ROI. If modern Wall Street had its way, we'd be parking in Goofy 47, then riding a bus to the entrance instead of a monorail. After all, it would reduce costs by 2.7 cents per guest per day!
Well... except for the fact that it would have made the Magic Kingdom a lot less "magical" to kids like us, and we would have thrown a fit and begged to go to Bush Gardens instead (yeah, I'll admit it... my parents could have easily saved a few hundred bucks each summer by skipping the Magic Kingdom, getting a room at the Contemporary Resort, and letting me ride the monorail all day with complete legitimacy as a hotel guest instead. But I digress...)
Even if you assume 100% of that $4k was capital construction cost, it's something that's going to have a useful life measured in centuries. Plus, much of that cost was self-inflicted by Verizon itself. If Verizon were smart, they would have pulled new fiber everywhere, left the copper in place, then spun off a new company to own the copper wires (but not the easements through which they ran) and sell landline phone service & discount DSL to poor people. The new company would have never been a long-term threat to Verizon (with no ROW of their own, and only an easement to run copper UTP, they'd have been hitting their head against the ceiling within a decade or two), and the value to a buyer would have been far more than the scrap value of the copper wires and lost customers that Verizon wants to ditch anyway.
Almost, but not quite. There are plenty of Android phones with bootloaders that are unlocked (officially or otherwise), but are still stuck with old kernels because they depend upon binary loadable kernel modules that are not themselves open-source. Remember, Linux doesn't have a stable ABI, so loadable kernel modules ("drivers", in Windows parlance) are specific to a kernel version.
This is probably the #1 source of recurring grief at xda-developers.com. Every new version of Android ships with a new kernel that breaks every binary driver (camera, GPS, wi-fi, sensors, bluetooth, etc) that came before it. IMHO, it's the #1 reason why Google needs to just fork Android's kernel (since 99.9% of the changes are Android-specific, and few mainline changes really matter to Android devices) and commit to having either a stable ABI, or at least come up with a reasonable thunking layer to enable binary modules that are a year or two old to continue working with newer kernels, absent some *really* good reason (like a catastrophically bad security vulnerability) to justify a compatibility-breaking change.
Some people might not like voiding their warranty the day they buy their phone.
Which is why we all make a nandroid backup before flashing a new firmware.
> Some people might not like voiding their warranty the day they buy their phone.
Manufacturers can lie about warranty-invalidation until they're blue in the face. The Magnuson-Moss Warranty Act ( http://en.wikipedia.org/wiki/Magnuson%E2%80%93Moss_Warranty_Act ) is a potent weapon that no manufacturer, not even Apple in their most arrogant AT&T-exclusive hissy fit, would dare to push back against because the FTC will smack them down and make a total example out of them.
Under Magnuson-Moss, a manufacturer can only deny warranty coverage if they can demonstrate that whatever the consumer did was literally the cause of the failure... and historically, the FTC hasn't made their job easy. They basically get one chance to make their case to the FTC, and if the FTC thinks the company is harassing the customer and wasting their time on a silly excuse, it will instantly smack them down and hit them with a huge fine.
If the manufacturer wanted to use "we had to reflash it via JTAG to stock" as an excuse for denying the claim or imposing a service fee, they'd have to testify that they don't routinely JTAG-reflash to stock as a troubleshooting step anyway.
If they tried to argue that you somehow triggered a condition via software that caused damage (say, setting a pair of directly-connected GPIO pins to outputs, with one high and one low), they'd still be backed up against the wall and told they were idiots for not putting a resistor between them, or at least going out of their way to make it abundantly clear to end users that custom firmware must never, ever do that specific action. In stark contrast to most consumer non-law, the FTC takes consumer rights seriously, and doesn't take crap from companies who try to wave vague disclaimers around and use them as an excuse and blanket license to run roughshod over consumers. The barrier isn't quite insurmountable, but a company that tried to fight it would have an uphill battle, and quickly discover that its usual dirty tricks weren't going to work this time around.
Companies doing dirty tricks with warranty coverage is nothing new. The same things phone manufacturers try to do today, American automakers did to our parents and grandparents openly and proudly, with a dash of extra salt to rub into consumer wounds ~30 years ago.
Magnuson-Moss is a rare gem of consumer-protection law passed by an angry congress fed up with the increasingly-bold abuses of the 3 most powerful companies in America at the time. Apple, Samsung, HTC, and Motorola might be powerful... but they're *nothing* compared to the "Big Three" American automakers circa 1975, and they know it.
Unfortunately, it's NOT against the law for a company to blatantly lie about its legal responsibilities, so companies can say anything and put all the restrictive text they want to put in their warranty descriptions. You just have to know that when push comes to shove, all you have to do is whisper the magic phrase "Magnuson-Moss" to get your complaint *instantly* escalated to the most senior manager on site and get total white-glove treatment and profuse apologies for the "misunderstanding" (inevitably blamed on the tier-1 support staff, who were just doing what the script told them to do).
It depends on the site.
Stackoverflow is by far the most tolerant of questions that don't fit neatly into a perfect box that can be wrapped and declared to be answered. Questions that deviate from the formula DO eventually get moderated in a way that allows them to fade from view and get pushed to the bottom of search results, but there's (fortunately) very little of the religious zeal for stamping out Badthink that seems to infect most of its other sites.
In contrast to SO, the other SE sites really seem to attract moderators who fall into the "ISTJ" personality type (strong sense of right & wrong, disdain for hypothetical problems or theoretical solutions compared to concrete solutions to clearly-defined problems, deference to authority and tradition, sense that the destination or goal is the whole reason for doing something, and almost zero tolerance for "inefficiency"). As luck would have it, this group tends to clash rather badly with the INTP (and weakly-J INTJ) personality types that are common among programmers, and StackOverflow is the Jerusalem where the two share an uneasy & tense coexistence... surrounded by a StackExchange ocean of more public and visible conflict(*).
By comparison, ServerFault is *hardcore* ISTJ territory, and the INTx-ISTJ conflict is extremely visible. Many SF mods act like it's their religious duty to stamp out even the slightest hint of something that might be interpreted as circumvention of official authority (or even just frowned upon). To an ISTJ, rules are meant to be followed and enforced... including blanket rules with awkward edge cases, if carving out exceptions for those edge cases would turn a rule that can be concisely expressed by a single sentence into 27 pages of flowcharts, BNF grammars, and appendices. In contrast, the INTx (particularly INTPs) will meticulously document the scenarios where the "one size fits nobody" rule produces absurd results, then go nuclear when the ISTJs say they don't care and it doesn't matter.
Ultimately, what SE in general needs is a Wikipedia-like "discussion" mechanism to move the discussions a step away from the factual essence of the main page, but nevertheless recognize that healthy disagreement and discussion can be valuable, especially for topics where yesterday's best-practice might be today's horrific security vulnerability or architectural dead end. The fact is, there are lots of problems that really *don't* have viable solutions (yet), and documentation of the specific roadblocks can be valuable in itself.
(*)For anyone who cares, the MBTI personality type most likely to clash with yours is NOT necessarily its polar opposite. The truly epic FIGHTS tend to occur between two individuals of the SAME introversion/extroversion type, but conflict with respect to the remaining 3 attributes. An INTP and INTJ with common interests might be new best friends for the first 10 minutes of a 4 hour car trip, and bitter enemies by the end.
That's a mostly accurate summary of the outwardly visible effects a nontechnical end user who only wants to "surf and talk" might see, but it doesn't quite describe the actual problem. If you know exactly what's going on behind the scenes, there ARE ways to do simultaneous voice + data on Sprint (though most of them require spending more money for thirdparty services and doing an end-run around Sprint itself).
Let's start with voice calls. If you want to terminate a voice call through Sprint, there's exactly one way it can happen: via circuit-switched CDMA. The call travels between Sprint and whomever is at the other end of the call over the public switched telephone network (PSTN). Between Sprint's switching center and the phone, the call is transported via CDMA2000 voice.
No Sprint phone I'm aware of can simultaneously handle an active CDMA voice call and IP data using either EVDO or 1xRTT. It's physically impossible due to the way both are implemented on Sprint's phones. HOWEVER, in most of Sprint's recent phones, there's no hardware reason why you can't have an active voice call AND use data via Wimax, LTE, and/or wifi. Some Sprint phones have shipped with firmware that disabled data use during voice calls, but that was mainly due to tech support and demographics. Basically, they didn't feel like dealing with less-technical users who couldn't be assumed to understand the difference between the different data modes, be aware of their connectivity from minute-to-minute, and actively manage their data connection mode in order to do both at once.
That said, there's nothing (besides battery life and added subscription cost) to stop you from acquiring VoIP service from someone besides Sprint, running a SIP client on your phone, and making simultaneous VoIP calls while using any working data mode, including EVDO and (maybe, bandwidth permitting) 1xRTT. You can even use call forwarding to forward your incoming Sprint calls to the VoIP number. The downside is that your phone will drain the battery a LOT faster, because you'll have to actively poll for incoming calls (normal incoming calls are handled by having the phone poll Sprint's towers via the same mechanism used for text messages; with VoIP, the phone is polling twice as much, and has to maintain an active data connection to do it). One compromise is to make your outgoing calls via VoIP, but let your incoming calls continue to come in through Sprint (losing EVDO/1xRTT data connectivity when it happens).
So... you might be wondering... if end users can get simultaneous voice+data, even over CDMA data modes, by using VoIP service acquired independently of Sprint... why can't Sprint itself do it? Basically, their switching equipment can't handle it. Nothing that couldn't mostly be hacked around, but it would have been expensive, likely to cause problems (witness the thrashing many wimax Sprint phones do when they can't make up their mind between EVDO and wimax, and just keep breaking the data connection and thrashing wildly between the two), and would have still left Sprint with compromises compared to HSPA+.
For what it's worth, these problems are nothing new... GSM networks went through the exact same hardware problem 10 years ago when they transitioned from TDMA-based GSM/GPRS/EDGE to WCDMA-based UMTS/HS(D|U)PA(+). The main difference is that European phone companies actually WERE able to buy off the shelf switching equipment to deal with it, whereas Sprint would have had to cobble its own half-baked solution in-house. Qualcomm WAS actually working on SVDO to replace EVDO (doing more or less the same thing as HSPA+), but most CDMA carriers in other countries decided to skip EVDO and just transition to UMTS/HS(D|U)PA(+) instead. Sprint and Verizon couldn't do that, because they didn't have the pair of 10MHz uplink and downlink channels in all of their markets.
Sprint AND Verizon together would have been enough to motivate Qualcomm to finish development of SVDO, but Verizon hoped to strike a deathblow against Sprint by formally aband
No, "Hide extensions of known file types (Recommended)" was the first shot in Microsoft's war against its customers.
Actually, it's more like the government forced lenders to drop their standards for certain favored groups and backed it up with federal insurance, then the lenders realized they could just drop their standards for everyone to the same level and wash their hands of its consequences. The same loopholes that allowed a struggling working-poor person to qualify for a home with zero down and an artfully-contrived assessment of his or her assets & annual income allowed middle-class wannabe-investors to buy pre-construction condos with zero down for double their realistic value in the hopes that they'd be able to flip them for 4 times their realistic value. This inflated prices in general, and got otherwise-sane middle-class buyers buying a home for themselves to agree to stretch their finances to the breaking point to afford to buy a house that everyone thought would become unaffordable to them forever if they didn't buy immediately. In cities like Miami, houses that would have barely gotten $80,000 in 1995 were bought sight-unseen for $250,000 in 2003, and $399,000 in 2005, by speculators planning to demolish them and build a million-dollar replacement to flip in their place. Caught in the middle was the example middle-class family facing the prospect being priced out of the housing market forever.
The only, and I mean *ONLY* thing that kept the market from going completely nuclear, were the guidelines that made it hard to qualify for a loan of more than ~$360k with zero down, and almost impossible to qualify for a loan of more than ~$460k without real proof that you could afford it and a major down payment. For a time, both were de-facto price caps in South Florida... tiny 650sf condos with no reserved parking were almost universally $360,000, two-bedroom townhomes (or single-family homes in borderline-ghetto neighborhoods) were almost universally $460k, and everything else involved $360k or $460k plus $50k-100k cash down. Without those limits, there would have absolutely been people buying zero-down million-dollar 1 bedroom condos, despite their annual incomes being less than just a year's worth of INTEREST payments on the loan.
> I am yet to see DNS fail badly. I have seen plenty of people who don't understand it say it does, when the problem is invariably routing or a firewall.
Note the key phrase, "who don't understand it" and its modifiers "routing or a firewall". There's a HELL of a lot of people who happen to fall into that category, and whose frustration goes off the scale when something fails to work because the slightest configuration problem will break it, and if you manage to avoid a subtle semantic bug in a zonefile somewhere, factors upstream that are beyond your control can still break it in ways that are almost impossible to distinguish from that same hypothetical zonefile bug. Bind is a cruel, heartless, sadistic, and demanding master. I struggle to think of anything that universally strikes fear in the heart of otherwise brave men than "Can you set up the DNS for us from scratch? We registered the domain yesterday. The computer's over there, and here's the Ubuntu installation DVD."
DNS only "just works" when some OTHER unfortunate soul has already set it up and spent the day troubleshooting it for you. Worst of all, there's a big, fuzzy gray area of "works for me, and apparently for him, but not for you for some unknown reason".
So, yes... for 17 or 18 artisan-level gurus who've achieved Englightenment, DNS is easy and straightforward to set up. For the other 99.9% of individuals unfortunate enough to find themselves tasked with the duty of setting it up at the server end, it's pure hell, and Bind was a punishment invented by God for sadistic entertainment purposes.
When I talk about "DNS Failure", I'm not talking about it from the perspective of pure end users who connect to a network and make use of DNS that somebody else has already gotten to work for them. I'm talking about the unrelenting hell of being someone who lacks control of his upstream network configuration trying to set up his own DNS server and make sense of errors that could be caused by just about anything, or (almost) nothing at all, and can take a relative eternity to troubleshoot when it happens.
You're right, but what's the pixel resolution of a "720p class" TV, and how big is its "19 inch class" screen?
The situation I'm complaining about is what it would be like if that 2x4 piece of wood were a "2x4 class board product" whose actual dimensions could be anywhere between 1 x 2.5 to 1.4 x 3.5, and endless racks of 70-cent "board products" at Home Depot made a real, honest to god 2x4 cost $15 and have to be special-ordered.
The problem is, DNS is like USB, and IPv4 is like RS232. If you're anywhere close to being right, you can probably get ipv4 (or a real serial port set to 9600-8-N-1) to work well enough to give you clues about what the real problem is. In contrast, DNS (like USB) tends to just fail hard and catastrophically, giving no obvious clues about what might actually be wrong.