> Then either don't allow calculators at all or provide standard calculators.
Or require students to use a specific model of calculator, with their names printed on the back. Before each test, collect the calculators, shuffle them, and hand them out randomly. Statistically, absent wholesale class-wide collusion, your problem is solved.
> Um, the contract that you agreed to that allows access to the carriers private data/voice network. Laid down by the same carrier that you are contracted with.
And, in the United States, if your carrier happens to be Verizon or AT&T, they're subject to the same consent decree that broke up AT&T's monopoly 25 years ago & forced them to allow consumers to own their own phones. At first, it didn't matter much, because all phones were... well... phones. But within a couple of years, phones started to pick up new features, some of which had absolutely nothing to do with being a "phone", and other devices that were never conceived as phones began to acquire the ability to act like phones.
> These are not standalone devices, and have different risks to whom you have a contract with, thus different rules.
Newsflash: a hacked DOCSIS3.0 cable modem can cause WAY more service disruption than the most hacked Android phone to have ever existed on planet earth. Yet, by law, you can walk into Best Buy (or some other store) and buy your own anyway, whether Comcast likes it or not. Cable modems are literally radio transmitters (and fairly powerful ones, at that), and their broadcasts share cable with customers over a shockingly large area that makes the area served by a single cell tower look small by comparison.
If the carriers want that much control and are that concerned about protecting the integrity of their networks, let them get together and define the specs for a mobile network interface that's basically a "black box" wireless network card having the approximate form factor of a thick SD card whose external connectors consist of power, RF, I2C, and ethernet. Then I can buy whatever pocket computer I like, stick their network interface into it, and then users and carriers can peacefully coexist on opposite sides of a well-defined wall of separation.
> Rooters will just get a different phone in the first place, like a Nexus 1 or something by HTC.
Actually, I'd argue the opposite is true. People who root and reflash tend to be fanatically loyal to their phone's maker compared to "normal" users. Why? Because people who root & reflash invested lots of time acquiring familiarity with their phone's architecture, to the point where Android phones by other makers seem to be ENORMOUSLY more different & alien than they do to "normal" users. Not to mention the fact that 95% of their casual daily online social lives probably revolve around others who are ALSO into rooting and customizing their phones. These aren't users who'll switch to Samsung or HTC because the bundled Twitter client has a pretty translucent blue background instead of a red one, and they prefer blue instead.
The truth is, if Motorola hadn't fucked up their later phones, nearly every single owner of a rooted Droid would have eventually bought a Motorola phone to replace it. If you read around the various sites, the users who are the most upset and screaming the loudest about this are people who own a Verizon Droid *right now*, because up to this point the idea of replacing it someday with a non-Motorola phone was borderline-unthinkable. In a very real way, they feel like they've been kicked in the balls and shat upon by someone they admired.
> A USB cable is a specialist tool when it comes to mobiles, most users have never, and probably will never, connected their phone to their computer.
Er, no. I'll grant that a JTAG is a specialist tool, even if plenty of Slashdot's users own one.
I'll even partly concede you might have a point if the device can be resuscitated using a DB25f connector, some wires, and access to a PC with a real parallel port (or a 2x13 IDC header on the mobo with the same signals, still quite common even on PCs without an actual parallel port on the back), just because getting *anything* that bitbangs a raw parallel port to work reliably under Windows is *harder* than rooting and reflashing most Android phones (thanks, Bill!).
But a USB cable? No. Every Android phone I know of includes one in the box, and anybody who's reflashed his phone to a custom ROM (or trying to do so) is going to be quite intimately familiar with its use by the time he gets to any point where bricking his phone is a real possibility.
No, a mortgage is a loan secured by the property. When you buy a phone, even a subsidized phone, you legally own it outright the moment you sign the credit card slip or hand the cashier the money.
If you went the small-claims court route, you'd really file a lawsuit against Verizon. They're the ones who ultimately sold the phone to you, and more importantly, they have a "nexus" in pretty much every state in America. Small-claims courts tend to be very, very unsympathetic towards large corporations seeking to change venue. Unless they can claim the court lacks jurisdiction because they don't have a legal nexus in the state, their odds of getting the trial moved are basically nonexistent. Not to mention that in the real world, it would cost more money to pay their law firm to file claims and responses seeking a venue change than it would cost to just refund the purchase price of your phone (or give you a new phone for free as a replacement).
The truth is, you don't sue over something like this in small claims court to get the money. At the end of the day, you'll lose a day of work and $50-100. The satisfaction comes from knowing that your little lawsuit ended up costing THEM a thousand dollars or more in legal fees and lost productivity. No company the size of Verizon or Motorola would ever DARE to not involve at least one lawyer to review the case, because they can't risk missing something serious and having penalties or sanctions blow up in their face.
A little good news -- the cost of a ghetto-fabulous JTAG programmer has come down. WAY down. FTDI released a single-chip USB host a few months ago with onboard JTAG capabilities AND ready to use royalty-free driver DLL for Windows. It's around $12 for the bare chip, and around $28 for a breakout board with the chip and support components soldered to it ( http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=768-1030-ND )
It still sucks to have to even screw with JTAG for this, but it's not *completely* hopeless.
Now, what would be REALLY funny is if someone motivated by desire to root & reflash his DroidX ends up using his new knowledge and JTAG to crack something whose security really DOES matter to Motorola, like his cable box. The Law of Unintended Consequences has an amusing way of extracting periodic revenge;-)
> Oh, so we just have to wait a couple of weeks for some teenager to crack it. Awesome.
If it's hardware-based 2048-bit encryption, it's not going to be hacked in a couple of weeks. CSS encryption used on DVDs was a complete joke of an encryption standard, and everyone knew it at the time. It was security by obscurity, and almost worked for a couple of years.
When it comes to military-grade hardware encryption, Motorola doesn't screw around.
If Motorola doesn't back down, realistically the best prospect for reflashing a DroidX is a homebrew JTAG programmer built with a FTDI 2232H on a breakout board -- http://ftdichip.com/Products/FT2232H.htm
As painful as it is to even imply that I might defend Verizon against anything, this is Motorola's doing first and foremost. I'm sure Verizon isn't *DISPLEASED* by it, but it actually goes straight to Motorola's corporate culture. They were making Linux-based phones locked down tighter than Tori Speling's 90210 chastity belt when the iPhone was just the wet dream of Steve Jobs. Their first Android phone, the Droid, wasn't locked. Their second, the Milestone -- sold unlocked, to end users in Europe, at full retail -- WAS locked.
On the brighter side, if the fact this has been on Slashdot for a few hours and is almost up to 7 or 8 pages by now indicates anything, early sales of the DroidX aren't likely to make Verizon happy, and might very well induce some Verizon execs to make an angry phone call or two over to Motorola and demand a new bootloader and public apology *yesterday*. The *last* thing Verizon wants right now is a "hold different" debacle of their own, especially with Sprint selling Evos as fast as they can and the Epic4g around the corner as soon as the Evo's exclusivity period runs out in a few more days.
You might want to check out the Samsung Galaxy (coming to VZ sometime soon). HTC isn't entirely free of sin, either. Read all the complaints about them at xda-developers.com over their kernel source. Basically, the GPL2 requires source for anything that's compiled into a monolithic kernel, but Linus (and most reasonable people) have come up with a tolerable compromise for proprietary binary modules that involves neatly packaging them up as loadable kernel modules with a well-defined interface so you can rebuild the rest of the kernel around it without losing whatever hardware support is provided by those proprietary binary modules. The problem is, HTC compiled their drivers straight into the kernel instead of making them proper loadable kernel modules, then simply hacked them out of the source before releasing it. The net result is that you can't use the source HTC released to build a functional Android-ready Linux kernel for their phones, and we're STILL reduced to ripping rom images from newer phones in the same family in a desperate effort to get them to work on our phones.
I don't know whether Samsung does the same thing, or whether THEY release their binaries as proper.ko modules. I'm keeping my fingers crossed for now (planning to get an Epic4g myself, assuming something like this doesn't crop up at the last second to ruin it). I'm still trying to verify it myself, but if they do, then Samsung deserves major praise (if only because the rest of the handset makers are so completely fucked).
Yes they do. Did you miss last month's week of Android Outrage when AT&T released their new phones officially locked to prevent you from installing apps obtained from anywhere besides their app store, and the wails when HTC pulled the sync app that allowed users to get around that restriction?
I'm pretty sure that in the US, the car's electrical system and ECU is required by law to basically commit suicide when the airbags deploy. I've been told that if your car gets into any accident where the airbags deploy, it's going to be declared a total loss by the insurance company unless it's literally a Porsche or Jaguar and worth spending more than $20,000 to repair.
Because in America, an Openmoko is a nearly-useless GPRS paperweight. Ditto for Trolltech GreenPhone, and every other "open" phone. They couldn't do EDGE (in the US, lack of EDGE is a nearly automatic deal-breaker, because even today, few AT&T or T-Mobile customers can truly depend on UMTS working everywhere they go), let alone 1700/2100 or 850/1900 UMTS. IMHO, it's a major reason those phones have all largely failed.
Their argument that they were "development" phones not intended for real use was stupid. Very, very few businesses are going to buy a relatively expensive phone for pure abstract "development", and no slightly geeky elite user is going to spend that much money on a phone that's borderline useless as a phone ready to be tweaked, extended, and made MORE cool.
The US isn't the entire world market, but it's a big, important chunk of it, and any project that effectively writes it off completely is probably doomed from birth. The same is probably true of any American project that completely ignores Europe. Just look at Palm -- they released the Pre for Sprint while the rest of the world was literally begging for it. At the VERY least, they should have released a 1900/2100 UMTS Pre simultaneously with the CDMA Pre. Someday, companies in the US and Europe will realize that it's worth making at least a half-assed effort to ensure that anything that works in one place will at least have a chance of working in the other, because it's cheaper to spend a little more and end up with one device that works in both places than two devices that can independently flop and leave the vendor with useless, expensive inventory.
Google was smart -- they made sure the Nexus one would work fully in both Europe AND the US. Admittedly, T-Mobile wasn't the greatest choice of American networks, but their only real alternative would have been to have released it with 850MHz support from the start (frankly, I'm shocked they didn't), or go with a (still slightly experimental, apparently) tri-mode chipset capable of GSM800/900/1800/1900, UMTS850/1700/1900/2100, and CDMA2000@800/1800/1900 so it would work on Verizon & CDMA carriers in Latin America, Australia, New Zealand, Korea, and China.
Not true. When you buy the phone, liquidated damages (something called an "Early Termination Fee") gets tacked on if you don't complete the contract. It's your property the moment the credit card gets swiped through the reader or the cash goes in the register.
By law, if you request that your phone's SIM-lock (if GSM) be removed, or that you be given its MSL code (if it's CDMA), the phone company MUST give it to you as long as 30 days have elapsed since purchase. I'm not 100% sure, but I think even the 30-day waiting period can be eliminated if you waive your right to cancel the plan or return the phone.
American phone companies (at least Verizon and AT&T by virtue of being AT&T's offspring) aren't allowed to keep the phones as secured assets or lease them due to the consent decree that broke up AT&T's monopoly 25 years ago that prohibited them from forcing customers to lease phones instead of purchase them from independent sources on their own. I'm not sure, but I think the FCC incorporated its terms directly into its own regulations, so they probably apply to Sprint & T-Mobile as well. On the other hand, that might be the reason why Verizon was grudgingly forced to open its network to any phone you can physically figure out how to make work, while Sprint can get away with refusing to let anyone use any phone not purchased from Sprint.
I believe the first cell phone companies tried to lease phones to customers, but were prohibited from doing so by the FCC out of concern that if carriers were allowed to lease phones, the price of purchased phones would be wildly inflated and customers would be forced into leasing anyway. As a practical matter, subsidies turned out the same way (in the US, at least, though Google's fought the hard fight to at least try and change it a little).
OK, here's the two hundred ton elephant wearing a pink tutu dancing between Lady Gaga and Madonna that (surprisingly) nobody seems to have mentioned yet: for all intents and purposes, Symbian doesn't exist in the United States. As far as I know, you can't go to a store operated by Sprint, Verizon, AT&T, or T-Mobile and buy a brand new phone subsidized by the carrier that runs Symbian (maybe, MAYBE Nextel might have one imported from Japan, but I wouldn't count on it).
Actually, it goes deeper than that -- as far as I know, you can't even buy a phone running Symbian, period, that's capable of 3G data on any network in the United States (with the *possible* exception of an imported Japanese phone that by some miracle of God might work on Nextel). For whatever reason, Symbian is almost a synonym for "Expensive GSM phone that nevertheless can't do EDGE, and is capable of 3G UMTS only at 1900/2100MHz". Thus, no sane American likely to be remotely interested in a phone running Symbian is going to go out and spend $500 or more to buy an unlocked phone that's basically a GPRS paperweight capable of making voice calls in a pinch.
"Invisible and Irrelevant in America" == "Invisible and Irrelevant to American Journalists" (who happen to generate most of the English-language content that gets read worldwide, and highly influence the rest of it). Thus, daily headlines about iPhone and Android. Occasional mentions of Palm. <tongue location="cheek">Symbian? Is that, like, the new name for Palm or Windows Mobile or something? </tongue>
The fact that Symbian started enforcing code-signing a couple of years ago (effectively shutting out casual developers who've always been welcomed with open arms by Android and pre-Kin/7 Microsoft) certainly hasn't helped, either... the moment they did that, they effectively wrote off a big chunk of their most influential and outspoken EUROPEAN former users, too.
Apparently, copper is now SO expensive, there ARE companies making CCA (copper-coated aluminum) wire, too. It's still pretty rare, but it doesn't surprise me. Once word starts getting out about testing cable with a magnet, it's almost natural that the same Chinese companies would come up with a way to substitute a different metal that's still cheap compared to copper, but passes "the magnet test". In fact, here's just one company that makes all of the above: http://hengdeli.en.ecplaza.net/2.asp (used to prove that such companies exist, not encourage the use of their products). Though I guess that all things considered, CCA is at least a tiny improvement over CCS... at least aluminum conducts electricity with some degree of usefulness compared to steel, and doesn't rust through (I think it only oxidizes on the surface... not good, but at least the wire maintains its mechanical integrity instead of turning into red powder over time).
Still... is copper *really* that expensive now, or is the market for aluminum and copper in China just artificially distorted enough to tip the scale enough to make a difference?
I saw an Amiga 2500UX once. It was a demo unit owned by Creative Equipment in Miami, but from what I remember, there really wasn't much you could actually *do* with it at the time unless you were a college professor or grad student with a Unix-related obsession. There wasn't really any commercial software for it, and I doubt whether it even shipped with the necessary libraries to build anything more ambitious than maybe "Adventure", "Spacewar", and "Life" -- all of which were undoubtedly cool in the 70s, but didn't seem very interesting compared to "Bard's Tale", "Federation of Free Traders", and "Lemmings";-)
There's one tiny detail that throws a monkey wrench into just about *everyone's* understanding: a disturbingly large percentage of recent-vintage low-cost cable from China has been made using "copper-coated steel" (CCS) and NOT pure copper.
CCS has been used for years as the center conductor in 75-ohm coax, but its appearance in things like network cable alleged to be "cat5e" is a VERY recent phenomenon (as in, even 2 or 3 years ago, it was basically unheard of). When you throw CCS into the cable equation, everything you know about cable based upon past experience and external observances goes to hell. For the most part, it's safe to say that CCS does absolutely nothing GOOD for applications like ethernet, and has plenty of potential to do really bad things to it. I'd be shocked if the CCS cable pawned off as "cat5e" on eBay (and quite a few discount vendors online) would meet official cat5e specifications at a HUNDRED feet, let alone a thousand or more. Worst of all, unless you're making a point of watching out for CCS network cable, it actually looks BETTER upon casual inspection than decent pure-copper cable, because the wires are thicker. The problem is, the steel core does nothing for the signal, so you're basically trading a hair-thin AWG26 copper wire for a micron-thick hollow tube of copper electroplated onto a strand of steel wire, and using it in a scenario where the conductivity difference really, truly DOES matter.
Don't believe me? Go to eBay and search for "cat5e CCS" (sans quotes), checking the box to search the description as well. You'll find at least a page of results, and when you read the descriptions, you'll see that they most certainly ARE "CCS".
I'm not 100% sure, but I think the reason it's halfway-legal to sell CCS UTP cable as "cat5e Cable" is because (in America, at least), "cat5e" only has specific legal meaning if you use it in conjunction with "TIA/EIA-568-B", or make specific claims about its suitability for network use at a given speed and/or length. As long as you claim nothing beyond "AWG24 Cat5e Cable", you could probably get away with just about anything capable of conducting electricity and 4 twisted pairs of wire.
Anyway, beware. CCS "cat5e" cable is real, and is a growing problem unless you make a point of trying to avoid buying it.
I remember taking a class in "C and Unix" in college circa 1991 with a professor who was quite brilliant, but had a very, very thick Chinese (Cantonese, I believe... he mentioned living in Hong Kong quite a few times) accent and had a VERY hard time with the "L" sound.
I went through two entire classes taking notes on the use of "Unix share variables" before realizing that he meant "Unix SHELL variables"
He definitely didn't have a problem with "R" sounds, though. I distinctly remember him mentioning what was then still a strange, (relatively) new language called "C pruss pruss". Nowhere close to being a hard "American" 'r' sound, but unambiguously rhotic nonetheless.
---
(*)he didn't like chalkboards or overhead projectors, so we never got to see it in actual writing. Because "Unix" was a relatively minor part of the class and C was all I really cared about at the time, I didn't feel like burning $49 on the 'Unix' textbook just for the last 2 weeks of class. Hey, it was 1990, Amigas rocked, and Unix was something you suffered with on boring green VT100 terminals if you were too poor to afford a real computer with civilized IDE. Yeah, even in the stone age of ~1990, we had graphical, windowed IDEs with syntax highlighting and primitive context-sensitive interactive function lookups & argument references. At least, if you owned an Amiga, Mac, or Atari ST. I think PC users were still stuck in their 16-color 80-column textmode ghetto, even if they COULD use a mouse to move the blinking cursor block around the screen;-)
No, the real problem is that new phones come with ridiculously-undersized batteries. My HTC Hero's 1500mAH stock battery was a cruel joke that would be dead by mid-afternoon unless I used a task killer to kill Google Maps after arriving at the office in the morning. After I replaced it with a nice, beefy 3500mAH Seidio extended-life battery, my problems were solved. Now, the only thing that nukes the battery is using my phone as a ghetto wireless access point (after 2 or 3 hours of active wi-fi tethering, it almost gets hot enough to cook on, and the battery slowly drains even with the charger attached. But that's the only exception I've found so far, and I suspect it *might* be the result of a bug that's needlessly jacking the phone's wi-fi power output to max).
So... HTC releases the shiny, fast new Evo... and how big of a battery do they put in it? Another pathetic 1500mAH joke that's barely big enough to keep the phone in standby for 24 hours, or make it through a normal workday with periodic internet access and a few voice calls over the course of 6-8 hours.
Notice the disclaimer near the very bottom: "Not all services available *on* 4G". Notice they aren't saying "4G service isn't available everywhere"... they're saying "not all services can be provided via 4G". Besides data, what other service is there? Right... voice. That innocent-looking disclaimer is Sprint's get out of jail free card.
> with Sprint's EVO, you can do voice over the old network at the same time you do data over WiMax, > assuming you're in an area with WiMax at the time.
Are you sure? As I understand it, the problem is that the Evo was value-engineered to have a single radio module with two metaphorical "channels" (not channels in the FM sense, but channels in the code/time-division sense). CDMA2000 voice needs one channel. CDMA2000 EV-DO needs two channels. WiMax needs two channels by default, but can limp along with a single channel *if* (and *only* if) the tower's RF hardware explicitly supports single-channel mode. Apparently, the RF hardware deployed to basically all of Sprint/Clear's existing WiMax sites doesn't.
From what I understand, Sprint basically took the Qualcomm chipset used to enable CDMA2000 voice and UMTS data on the same phone (ie, Telus in Canada), and had them amend the WiMax standard so it would work with the same radio design. Ultimately, they ended up with something that wasn't quite completely alien to the original WiMax standard, but wasn't exactly 100% compatible with the original standard, either. So... when an Evo tries to establish a single-channel connection with a legacy WiMax tower, the tower thinks it's seeing a valid WiMax device with fatal interference and noise problems.
The net result is that if you live in a city with extensive legacy Clearwire 4G WiMax service today, simultaneous voice+data is unlikely to work for at least another year or two, because they aren't terribly eager to scrap millions of dollars worth of existing, barely-used infrastructure so we can take voice calls while tethered. On the other hand, if you live in a city that's due to get WiMax service soon, it might support simultaneous voice+data, depending upon when the hardware going into the hut next to the tower was actually purchased.
The big one, IMHO. Especially since Amazon is going to officially support Android Kindle readers. The fatal flaw of current dedicated ebook readers like the Kindle for things like programming books is the latency. Programming books are rarely read from start to finish like a novel, one page at a time. You flip around, and often jump between a couple of pages. e-Ink is totally unsuited for that particular use case. On the other hand, an Android tablet with 1280x720 display is big enough to display two Manning/OReilly-sized/formatted pages side by side and read comfortably (ok, slightly higher resolution would be nice, but I've read PDFs on a 19" 720p TV with VGA input and easily read side-by-side pages, so I know it's viable).
The problem is, we need a tablet that actually HAS 1280x720. So far, all the ones that actually seem available to buy in America are stuck in the 800/854x480 ghetto, which is kind of pointless if you already have a Nexus One/Droid/Evo/Incredible-class phone. With a little luck, they'll be abundant by Christmas:-)
Now, if there's a god, Amazon will release an Android Kindle API that lets developers create their own user interface for it (treating the DRM-protected part as a rectangular black box that can be embedded as a View into any Android app, with the API providing access to things like document metadata, telling it to turn to page 832, zoom to 60%, and set the upper-right origin of the page to the point 3.425% down and 5.015% over). It would be great for Amazon, too, because it'll give them access to free user interface R&D. 99% of the Android KindleReader apps will utterly suck, and another.9% will be flawed... but that remaining.1% will give Amazon some new UI ideas for its own hardware kindles that it can freely use.
It's easy... pair your iPad with just about any Android phone known to exist. Any rooted Android ROM worth mentioning can emulate a wireless access point, and there are lots of apps to do the same thing that can be downloaded and installed on any Android phone, rooted or not.
From the Android side, it's amusing that the first question journalists and users from the Apple side of the universe ask is whether a given tablet "supports 3G". Android owners don't care, because it'll be a cold day in hell before we pay our carrier yet more money for dedicated data service for a tablet. We just take for granted that any tablet we buy is going to wirelessly tether to our phone, and take advantage of the data service we already paid for. I haven't looked yet, but I'm sure someone has already written an app to let Android tablets make use of their Android phone's location services, too.
If you really want to blame anyone, blame the handset manufacturers determined to ram button-free plastic slabs down everyone's throats. It's one thing to hold a phone in one hand and navigate its keypad with your thumb by touch, maybe stealing a quick glance at the display for a fraction of a second once or twice in a minute. It's another matter entirely to try interacting with a tiny button-free plastic brick that somehow manages to ignore your intentional gestures, yet instantly reacts to even the tiniest accidental contact... usually, in ways that are even harder to abort once triggered(*). The real hazards aren't the people with their thumb over the keypad ready to press '1' for English, and '4' to delete the message... it's the people forced to take BOTH HANDS off the wheel and devote their full visual attention to the picture of a keypad so they can interact with their phone in even the most trivial way.
(*)You know... it's 3am, somehow someone's number gets activated in the phone app, and you accidentally touch the 'send' key. At that point, you can hit 'end' like a madman, hold it down, or do just about anything short of yanking out the battery, and it won't matter... the call will go through 2-3 seconds later, and they'll be mad at you for waking them up. It's not that the call physically can't be aborted... it's that the UI designers never bothered to accommodate the use case of an accidental call-initiation, so once you trigger the call, the UI just goes into a busy-wait until the attempt either succeeds or fails. I know, because 10 years ago, you COULD hit 'end' immediately after hitting 'send', and it worked. This specific problem emerged with the first PalmOS phones, became enormously worse under Windows Mobile, and has stayed equally bad under both Android and iPhone. It's a problem that's almost uniquely endemic to phones with "Glass UIs" that exists even on phones with hardkeys for send/end, and just gets even worse on phones with virtual send/end keys. OK, it might just be a "Sprint" thing, but I've heard enough complaints from others to believe it's really just a fundamental flaw in the way modern phone user interfaces are designed to work.
Slashdot Mirror
> Then either don't allow calculators at all or provide standard calculators.
Or require students to use a specific model of calculator, with their names printed on the back. Before each test, collect the calculators, shuffle them, and hand them out randomly. Statistically, absent wholesale class-wide collusion, your problem is solved.
> Um, the contract that you agreed to that allows access to the carriers private data/voice network. Laid down by the same carrier that you are contracted with.
And, in the United States, if your carrier happens to be Verizon or AT&T, they're subject to the same consent decree that broke up AT&T's monopoly 25 years ago & forced them to allow consumers to own their own phones. At first, it didn't matter much, because all phones were... well... phones. But within a couple of years, phones started to pick up new features, some of which had absolutely nothing to do with being a "phone", and other devices that were never conceived as phones began to acquire the ability to act like phones.
> These are not standalone devices, and have different risks to whom you have a contract with, thus different rules.
Newsflash: a hacked DOCSIS3.0 cable modem can cause WAY more service disruption than the most hacked Android phone to have ever existed on planet earth. Yet, by law, you can walk into Best Buy (or some other store) and buy your own anyway, whether Comcast likes it or not. Cable modems are literally radio transmitters (and fairly powerful ones, at that), and their broadcasts share cable with customers over a shockingly large area that makes the area served by a single cell tower look small by comparison.
If the carriers want that much control and are that concerned about protecting the integrity of their networks, let them get together and define the specs for a mobile network interface that's basically a "black box" wireless network card having the approximate form factor of a thick SD card whose external connectors consist of power, RF, I2C, and ethernet. Then I can buy whatever pocket computer I like, stick their network interface into it, and then users and carriers can peacefully coexist on opposite sides of a well-defined wall of separation.
> Rooters will just get a different phone in the first place, like a Nexus 1 or something by HTC.
Actually, I'd argue the opposite is true. People who root and reflash tend to be fanatically loyal to their phone's maker compared to "normal" users. Why? Because people who root & reflash invested lots of time acquiring familiarity with their phone's architecture, to the point where Android phones by other makers seem to be ENORMOUSLY more different & alien than they do to "normal" users. Not to mention the fact that 95% of their casual daily online social lives probably revolve around others who are ALSO into rooting and customizing their phones. These aren't users who'll switch to Samsung or HTC because the bundled Twitter client has a pretty translucent blue background instead of a red one, and they prefer blue instead.
The truth is, if Motorola hadn't fucked up their later phones, nearly every single owner of a rooted Droid would have eventually bought a Motorola phone to replace it. If you read around the various sites, the users who are the most upset and screaming the loudest about this are people who own a Verizon Droid *right now*, because up to this point the idea of replacing it someday with a non-Motorola phone was borderline-unthinkable. In a very real way, they feel like they've been kicked in the balls and shat upon by someone they admired.
> A USB cable is a specialist tool when it comes to mobiles, most users have never, and probably will never, connected their phone to their computer.
Er, no. I'll grant that a JTAG is a specialist tool, even if plenty of Slashdot's users own one.
I'll even partly concede you might have a point if the device can be resuscitated using a DB25f connector, some wires, and access to a PC with a real parallel port (or a 2x13 IDC header on the mobo with the same signals, still quite common even on PCs without an actual parallel port on the back), just because getting *anything* that bitbangs a raw parallel port to work reliably under Windows is *harder* than rooting and reflashing most Android phones (thanks, Bill!).
But a USB cable? No. Every Android phone I know of includes one in the box, and anybody who's reflashed his phone to a custom ROM (or trying to do so) is going to be quite intimately familiar with its use by the time he gets to any point where bricking his phone is a real possibility.
No, a mortgage is a loan secured by the property. When you buy a phone, even a subsidized phone, you legally own it outright the moment you sign the credit card slip or hand the cashier the money.
If you went the small-claims court route, you'd really file a lawsuit against Verizon. They're the ones who ultimately sold the phone to you, and more importantly, they have a "nexus" in pretty much every state in America. Small-claims courts tend to be very, very unsympathetic towards large corporations seeking to change venue. Unless they can claim the court lacks jurisdiction because they don't have a legal nexus in the state, their odds of getting the trial moved are basically nonexistent. Not to mention that in the real world, it would cost more money to pay their law firm to file claims and responses seeking a venue change than it would cost to just refund the purchase price of your phone (or give you a new phone for free as a replacement).
The truth is, you don't sue over something like this in small claims court to get the money. At the end of the day, you'll lose a day of work and $50-100. The satisfaction comes from knowing that your little lawsuit ended up costing THEM a thousand dollars or more in legal fees and lost productivity. No company the size of Verizon or Motorola would ever DARE to not involve at least one lawyer to review the case, because they can't risk missing something serious and having penalties or sanctions blow up in their face.
A little good news -- the cost of a ghetto-fabulous JTAG programmer has come down. WAY down. FTDI released a single-chip USB host a few months ago with onboard JTAG capabilities AND ready to use royalty-free driver DLL for Windows. It's around $12 for the bare chip, and around $28 for a breakout board with the chip and support components soldered to it ( http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=768-1030-ND )
It still sucks to have to even screw with JTAG for this, but it's not *completely* hopeless.
Now, what would be REALLY funny is if someone motivated by desire to root & reflash his DroidX ends up using his new knowledge and JTAG to crack something whose security really DOES matter to Motorola, like his cable box. The Law of Unintended Consequences has an amusing way of extracting periodic revenge ;-)
> Oh, so we just have to wait a couple of weeks for some teenager to crack it. Awesome.
If it's hardware-based 2048-bit encryption, it's not going to be hacked in a couple of weeks. CSS encryption used on DVDs was a complete joke of an encryption standard, and everyone knew it at the time. It was security by obscurity, and almost worked for a couple of years.
When it comes to military-grade hardware encryption, Motorola doesn't screw around.
If Motorola doesn't back down, realistically the best prospect for reflashing a DroidX is a homebrew JTAG programmer built with a FTDI 2232H on a breakout board -- http://ftdichip.com/Products/FT2232H.htm
About $30 worth of parts (plus shipping) from Digikey -- http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=768-1030-ND
God, talk about the Law of Unintended Consequences -- savor the irony if Motorola becomes the company that makes JTAG a household word :D
As painful as it is to even imply that I might defend Verizon against anything, this is Motorola's doing first and foremost. I'm sure Verizon isn't *DISPLEASED* by it, but it actually goes straight to Motorola's corporate culture. They were making Linux-based phones locked down tighter than Tori Speling's 90210 chastity belt when the iPhone was just the wet dream of Steve Jobs. Their first Android phone, the Droid, wasn't locked. Their second, the Milestone -- sold unlocked, to end users in Europe, at full retail -- WAS locked.
On the brighter side, if the fact this has been on Slashdot for a few hours and is almost up to 7 or 8 pages by now indicates anything, early sales of the DroidX aren't likely to make Verizon happy, and might very well induce some Verizon execs to make an angry phone call or two over to Motorola and demand a new bootloader and public apology *yesterday*. The *last* thing Verizon wants right now is a "hold different" debacle of their own, especially with Sprint selling Evos as fast as they can and the Epic4g around the corner as soon as the Evo's exclusivity period runs out in a few more days.
You might want to check out the Samsung Galaxy (coming to VZ sometime soon). HTC isn't entirely free of sin, either. Read all the complaints about them at xda-developers.com over their kernel source. Basically, the GPL2 requires source for anything that's compiled into a monolithic kernel, but Linus (and most reasonable people) have come up with a tolerable compromise for proprietary binary modules that involves neatly packaging them up as loadable kernel modules with a well-defined interface so you can rebuild the rest of the kernel around it without losing whatever hardware support is provided by those proprietary binary modules. The problem is, HTC compiled their drivers straight into the kernel instead of making them proper loadable kernel modules, then simply hacked them out of the source before releasing it. The net result is that you can't use the source HTC released to build a functional Android-ready Linux kernel for their phones, and we're STILL reduced to ripping rom images from newer phones in the same family in a desperate effort to get them to work on our phones.
I don't know whether Samsung does the same thing, or whether THEY release their binaries as proper .ko modules. I'm keeping my fingers crossed for now (planning to get an Epic4g myself, assuming something like this doesn't crop up at the last second to ruin it). I'm still trying to verify it myself, but if they do, then Samsung deserves major praise (if only because the rest of the handset makers are so completely fucked).
> AT&T doesn't cripple their phones
Yes they do. Did you miss last month's week of Android Outrage when AT&T released their new phones officially locked to prevent you from installing apps obtained from anywhere besides their app store, and the wails when HTC pulled the sync app that allowed users to get around that restriction?
I'm pretty sure that in the US, the car's electrical system and ECU is required by law to basically commit suicide when the airbags deploy. I've been told that if your car gets into any accident where the airbags deploy, it's going to be declared a total loss by the insurance company unless it's literally a Porsche or Jaguar and worth spending more than $20,000 to repair.
> Why not more love for projects like openmoko?
Because in America, an Openmoko is a nearly-useless GPRS paperweight. Ditto for Trolltech GreenPhone, and every other "open" phone. They couldn't do EDGE (in the US, lack of EDGE is a nearly automatic deal-breaker, because even today, few AT&T or T-Mobile customers can truly depend on UMTS working everywhere they go), let alone 1700/2100 or 850/1900 UMTS. IMHO, it's a major reason those phones have all largely failed.
Their argument that they were "development" phones not intended for real use was stupid. Very, very few businesses are going to buy a relatively expensive phone for pure abstract "development", and no slightly geeky elite user is going to spend that much money on a phone that's borderline useless as a phone ready to be tweaked, extended, and made MORE cool.
The US isn't the entire world market, but it's a big, important chunk of it, and any project that effectively writes it off completely is probably doomed from birth. The same is probably true of any American project that completely ignores Europe. Just look at Palm -- they released the Pre for Sprint while the rest of the world was literally begging for it. At the VERY least, they should have released a 1900/2100 UMTS Pre simultaneously with the CDMA Pre. Someday, companies in the US and Europe will realize that it's worth making at least a half-assed effort to ensure that anything that works in one place will at least have a chance of working in the other, because it's cheaper to spend a little more and end up with one device that works in both places than two devices that can independently flop and leave the vendor with useless, expensive inventory.
Google was smart -- they made sure the Nexus one would work fully in both Europe AND the US. Admittedly, T-Mobile wasn't the greatest choice of American networks, but their only real alternative would have been to have released it with 850MHz support from the start (frankly, I'm shocked they didn't), or go with a (still slightly experimental, apparently) tri-mode chipset capable of GSM800/900/1800/1900, UMTS850/1700/1900/2100, and CDMA2000@800/1800/1900 so it would work on Verizon & CDMA carriers in Latin America, Australia, New Zealand, Korea, and China.
Not true. When you buy the phone, liquidated damages (something called an "Early Termination Fee") gets tacked on if you don't complete the contract. It's your property the moment the credit card gets swiped through the reader or the cash goes in the register.
By law, if you request that your phone's SIM-lock (if GSM) be removed, or that you be given its MSL code (if it's CDMA), the phone company MUST give it to you as long as 30 days have elapsed since purchase. I'm not 100% sure, but I think even the 30-day waiting period can be eliminated if you waive your right to cancel the plan or return the phone.
American phone companies (at least Verizon and AT&T by virtue of being AT&T's offspring) aren't allowed to keep the phones as secured assets or lease them due to the consent decree that broke up AT&T's monopoly 25 years ago that prohibited them from forcing customers to lease phones instead of purchase them from independent sources on their own. I'm not sure, but I think the FCC incorporated its terms directly into its own regulations, so they probably apply to Sprint & T-Mobile as well. On the other hand, that might be the reason why Verizon was grudgingly forced to open its network to any phone you can physically figure out how to make work, while Sprint can get away with refusing to let anyone use any phone not purchased from Sprint.
I believe the first cell phone companies tried to lease phones to customers, but were prohibited from doing so by the FCC out of concern that if carriers were allowed to lease phones, the price of purchased phones would be wildly inflated and customers would be forced into leasing anyway. As a practical matter, subsidies turned out the same way (in the US, at least, though Google's fought the hard fight to at least try and change it a little).
OK, here's the two hundred ton elephant wearing a pink tutu dancing between Lady Gaga and Madonna that (surprisingly) nobody seems to have mentioned yet: for all intents and purposes, Symbian doesn't exist in the United States. As far as I know, you can't go to a store operated by Sprint, Verizon, AT&T, or T-Mobile and buy a brand new phone subsidized by the carrier that runs Symbian (maybe, MAYBE Nextel might have one imported from Japan, but I wouldn't count on it).
Actually, it goes deeper than that -- as far as I know, you can't even buy a phone running Symbian, period, that's capable of 3G data on any network in the United States (with the *possible* exception of an imported Japanese phone that by some miracle of God might work on Nextel). For whatever reason, Symbian is almost a synonym for "Expensive GSM phone that nevertheless can't do EDGE, and is capable of 3G UMTS only at 1900/2100MHz". Thus, no sane American likely to be remotely interested in a phone running Symbian is going to go out and spend $500 or more to buy an unlocked phone that's basically a GPRS paperweight capable of making voice calls in a pinch.
"Invisible and Irrelevant in America" == "Invisible and Irrelevant to American Journalists" (who happen to generate most of the English-language content that gets read worldwide, and highly influence the rest of it). Thus, daily headlines about iPhone and Android. Occasional mentions of Palm. <tongue location="cheek">Symbian? Is that, like, the new name for Palm or Windows Mobile or something? </tongue>
The fact that Symbian started enforcing code-signing a couple of years ago (effectively shutting out casual developers who've always been welcomed with open arms by Android and pre-Kin/7 Microsoft) certainly hasn't helped, either... the moment they did that, they effectively wrote off a big chunk of their most influential and outspoken EUROPEAN former users, too.
Apparently, copper is now SO expensive, there ARE companies making CCA (copper-coated aluminum) wire, too. It's still pretty rare, but it doesn't surprise me. Once word starts getting out about testing cable with a magnet, it's almost natural that the same Chinese companies would come up with a way to substitute a different metal that's still cheap compared to copper, but passes "the magnet test". In fact, here's just one company that makes all of the above: http://hengdeli.en.ecplaza.net/2.asp (used to prove that such companies exist, not encourage the use of their products). Though I guess that all things considered, CCA is at least a tiny improvement over CCS... at least aluminum conducts electricity with some degree of usefulness compared to steel, and doesn't rust through (I think it only oxidizes on the surface... not good, but at least the wire maintains its mechanical integrity instead of turning into red powder over time).
Still... is copper *really* that expensive now, or is the market for aluminum and copper in China just artificially distorted enough to tip the scale enough to make a difference?
I saw an Amiga 2500UX once. It was a demo unit owned by Creative Equipment in Miami, but from what I remember, there really wasn't much you could actually *do* with it at the time unless you were a college professor or grad student with a Unix-related obsession. There wasn't really any commercial software for it, and I doubt whether it even shipped with the necessary libraries to build anything more ambitious than maybe "Adventure", "Spacewar", and "Life" -- all of which were undoubtedly cool in the 70s, but didn't seem very interesting compared to "Bard's Tale", "Federation of Free Traders", and "Lemmings" ;-)
There's one tiny detail that throws a monkey wrench into just about *everyone's* understanding: a disturbingly large percentage of recent-vintage low-cost cable from China has been made using "copper-coated steel" (CCS) and NOT pure copper.
CCS has been used for years as the center conductor in 75-ohm coax, but its appearance in things like network cable alleged to be "cat5e" is a VERY recent phenomenon (as in, even 2 or 3 years ago, it was basically unheard of). When you throw CCS into the cable equation, everything you know about cable based upon past experience and external observances goes to hell. For the most part, it's safe to say that CCS does absolutely nothing GOOD for applications like ethernet, and has plenty of potential to do really bad things to it. I'd be shocked if the CCS cable pawned off as "cat5e" on eBay (and quite a few discount vendors online) would meet official cat5e specifications at a HUNDRED feet, let alone a thousand or more. Worst of all, unless you're making a point of watching out for CCS network cable, it actually looks BETTER upon casual inspection than decent pure-copper cable, because the wires are thicker. The problem is, the steel core does nothing for the signal, so you're basically trading a hair-thin AWG26 copper wire for a micron-thick hollow tube of copper electroplated onto a strand of steel wire, and using it in a scenario where the conductivity difference really, truly DOES matter.
Don't believe me? Go to eBay and search for "cat5e CCS" (sans quotes), checking the box to search the description as well. You'll find at least a page of results, and when you read the descriptions, you'll see that they most certainly ARE "CCS".
I'm not 100% sure, but I think the reason it's halfway-legal to sell CCS UTP cable as "cat5e Cable" is because (in America, at least), "cat5e" only has specific legal meaning if you use it in conjunction with "TIA/EIA-568-B", or make specific claims about its suitability for network use at a given speed and/or length. As long as you claim nothing beyond "AWG24 Cat5e Cable", you could probably get away with just about anything capable of conducting electricity and 4 twisted pairs of wire.
Anyway, beware. CCS "cat5e" cable is real, and is a growing problem unless you make a point of trying to avoid buying it.
I remember taking a class in "C and Unix" in college circa 1991 with a professor who was quite brilliant, but had a very, very thick Chinese (Cantonese, I believe... he mentioned living in Hong Kong quite a few times) accent and had a VERY hard time with the "L" sound.
I went through two entire classes taking notes on the use of "Unix share variables" before realizing that he meant "Unix SHELL variables"
He definitely didn't have a problem with "R" sounds, though. I distinctly remember him mentioning what was then still a strange, (relatively) new language called "C pruss pruss". Nowhere close to being a hard "American" 'r' sound, but unambiguously rhotic nonetheless.
---
(*)he didn't like chalkboards or overhead projectors, so we never got to see it in actual writing. Because "Unix" was a relatively minor part of the class and C was all I really cared about at the time, I didn't feel like burning $49 on the 'Unix' textbook just for the last 2 weeks of class. Hey, it was 1990, Amigas rocked, and Unix was something you suffered with on boring green VT100 terminals if you were too poor to afford a real computer with civilized IDE. Yeah, even in the stone age of ~1990, we had graphical, windowed IDEs with syntax highlighting and primitive context-sensitive interactive function lookups & argument references. At least, if you owned an Amiga, Mac, or Atari ST. I think PC users were still stuck in their 16-color 80-column textmode ghetto, even if they COULD use a mouse to move the blinking cursor block around the screen ;-)
No, the real problem is that new phones come with ridiculously-undersized batteries. My HTC Hero's 1500mAH stock battery was a cruel joke that would be dead by mid-afternoon unless I used a task killer to kill Google Maps after arriving at the office in the morning. After I replaced it with a nice, beefy 3500mAH Seidio extended-life battery, my problems were solved. Now, the only thing that nukes the battery is using my phone as a ghetto wireless access point (after 2 or 3 hours of active wi-fi tethering, it almost gets hot enough to cook on, and the battery slowly drains even with the charger attached. But that's the only exception I've found so far, and I suspect it *might* be the result of a bug that's needlessly jacking the phone's wi-fi power output to max).
So... HTC releases the shiny, fast new Evo... and how big of a battery do they put in it? Another pathetic 1500mAH joke that's barely big enough to keep the phone in standby for 24 hours, or make it through a normal workday with periodic internet access and a few voice calls over the course of 6-8 hours.
Notice the disclaimer near the very bottom: "Not all services available *on* 4G". Notice they aren't saying "4G service isn't available everywhere"... they're saying "not all services can be provided via 4G". Besides data, what other service is there? Right... voice. That innocent-looking disclaimer is Sprint's get out of jail free card.
> with Sprint's EVO, you can do voice over the old network at the same time you do data over WiMax,
> assuming you're in an area with WiMax at the time.
Are you sure? As I understand it, the problem is that the Evo was value-engineered to have a single radio module with two metaphorical "channels" (not channels in the FM sense, but channels in the code/time-division sense). CDMA2000 voice needs one channel. CDMA2000 EV-DO needs two channels. WiMax needs two channels by default, but can limp along with a single channel *if* (and *only* if) the tower's RF hardware explicitly supports single-channel mode. Apparently, the RF hardware deployed to basically all of Sprint/Clear's existing WiMax sites doesn't.
From what I understand, Sprint basically took the Qualcomm chipset used to enable CDMA2000 voice and UMTS data on the same phone (ie, Telus in Canada), and had them amend the WiMax standard so it would work with the same radio design. Ultimately, they ended up with something that wasn't quite completely alien to the original WiMax standard, but wasn't exactly 100% compatible with the original standard, either. So... when an Evo tries to establish a single-channel connection with a legacy WiMax tower, the tower thinks it's seeing a valid WiMax device with fatal interference and noise problems.
The net result is that if you live in a city with extensive legacy Clearwire 4G WiMax service today, simultaneous voice+data is unlikely to work for at least another year or two, because they aren't terribly eager to scrap millions of dollars worth of existing, barely-used infrastructure so we can take voice calls while tethered. On the other hand, if you live in a city that's due to get WiMax service soon, it might support simultaneous voice+data, depending upon when the hardware going into the hut next to the tower was actually purchased.
> Ebook reader.
The big one, IMHO. Especially since Amazon is going to officially support Android Kindle readers. The fatal flaw of current dedicated ebook readers like the Kindle for things like programming books is the latency. Programming books are rarely read from start to finish like a novel, one page at a time. You flip around, and often jump between a couple of pages. e-Ink is totally unsuited for that particular use case. On the other hand, an Android tablet with 1280x720 display is big enough to display two Manning/OReilly-sized/formatted pages side by side and read comfortably (ok, slightly higher resolution would be nice, but I've read PDFs on a 19" 720p TV with VGA input and easily read side-by-side pages, so I know it's viable).
The problem is, we need a tablet that actually HAS 1280x720. So far, all the ones that actually seem available to buy in America are stuck in the 800/854x480 ghetto, which is kind of pointless if you already have a Nexus One/Droid/Evo/Incredible-class phone. With a little luck, they'll be abundant by Christmas :-)
Now, if there's a god, Amazon will release an Android Kindle API that lets developers create their own user interface for it (treating the DRM-protected part as a rectangular black box that can be embedded as a View into any Android app, with the API providing access to things like document metadata, telling it to turn to page 832, zoom to 60%, and set the upper-right origin of the page to the point 3.425% down and 5.015% over). It would be great for Amazon, too, because it'll give them access to free user interface R&D. 99% of the Android KindleReader apps will utterly suck, and another .9% will be flawed... but that remaining .1% will give Amazon some new UI ideas for its own hardware kindles that it can freely use.
It's easy... pair your iPad with just about any Android phone known to exist. Any rooted Android ROM worth mentioning can emulate a wireless access point, and there are lots of apps to do the same thing that can be downloaded and installed on any Android phone, rooted or not.
From the Android side, it's amusing that the first question journalists and users from the Apple side of the universe ask is whether a given tablet "supports 3G". Android owners don't care, because it'll be a cold day in hell before we pay our carrier yet more money for dedicated data service for a tablet. We just take for granted that any tablet we buy is going to wirelessly tether to our phone, and take advantage of the data service we already paid for. I haven't looked yet, but I'm sure someone has already written an app to let Android tablets make use of their Android phone's location services, too.
If you really want to blame anyone, blame the handset manufacturers determined to ram button-free plastic slabs down everyone's throats. It's one thing to hold a phone in one hand and navigate its keypad with your thumb by touch, maybe stealing a quick glance at the display for a fraction of a second once or twice in a minute. It's another matter entirely to try interacting with a tiny button-free plastic brick that somehow manages to ignore your intentional gestures, yet instantly reacts to even the tiniest accidental contact... usually, in ways that are even harder to abort once triggered(*). The real hazards aren't the people with their thumb over the keypad ready to press '1' for English, and '4' to delete the message... it's the people forced to take BOTH HANDS off the wheel and devote their full visual attention to the picture of a keypad so they can interact with their phone in even the most trivial way.
(*)You know... it's 3am, somehow someone's number gets activated in the phone app, and you accidentally touch the 'send' key. At that point, you can hit 'end' like a madman, hold it down, or do just about anything short of yanking out the battery, and it won't matter... the call will go through 2-3 seconds later, and they'll be mad at you for waking them up. It's not that the call physically can't be aborted... it's that the UI designers never bothered to accommodate the use case of an accidental call-initiation, so once you trigger the call, the UI just goes into a busy-wait until the attempt either succeeds or fails. I know, because 10 years ago, you COULD hit 'end' immediately after hitting 'send', and it worked. This specific problem emerged with the first PalmOS phones, became enormously worse under Windows Mobile, and has stayed equally bad under both Android and iPhone. It's a problem that's almost uniquely endemic to phones with "Glass UIs" that exists even on phones with hardkeys for send/end, and just gets even worse on phones with virtual send/end keys. OK, it might just be a "Sprint" thing, but I've heard enough complaints from others to believe it's really just a fundamental flaw in the way modern phone user interfaces are designed to work.