Slashdot Mirror

NXP i.MX 8M Mini Applications Processor Evaluation Board

-40 C to +105 C good enough for you?

Also, this SoC: NXP Semiconductors MIMX8MM6DVTLZAA

From the Pine64 FAQ:

What is Pine A64 operating temperature?
The Pine A64 operating temperature qualified range from 0C to 70C.

You'll need active heating and cooling of some sort to go with a Pine64. I think NXP's i.MX line is more focused on what you're looking for.

Yes I do. When I first wondered if it would be possible to do such a thing I did some searching and found the following 2 posts from someone who had puttered around with it some:
post 1
post 2
My takeaway from that is that he was gong for more of a flare effect which he did sort of achieve but only with modifying the plate which would likely result in other legal issues. In the tests he did he was still operating in the 10s of watts range which isn't that much power. Instead my thought would be to throw the exposure off of the camera and massively underexpose the plate. To do this you would need to have a draw in the 100s of watts and have it be over a larger area near the plate. Also the LEDs to use looks to be these ones. Working in IR also has the advantage that many states have resigned their license plates so that they are easier to read in the IR spectrum as the cameras used for ALPRs are IR cameras. Also keep in mind that covering your license plate with anything may be illegal in your state as it is in Minnesota and that there may be vehicle illumination laws which apply but that doesn't appear to be the case in MN as they only cover visible light. But from what I can tell pumping out half a KW of IR in front of and behind your car in MN isn't illegal (IANAL) so long as the device doing it doesn't cover the license plate.

I used to love going to RadioShack and digging through the bins of components, looking for parts and inspiration... sadly that's no longer an option :(

Go to Digikey, and rummage through a bin with 6 million components to choose from.

I can believe that anyone is seriously arguing that hardware hacking was better in the "good ole' days".

Lightbulbs became electronics when they started to include switched mode power supplies.
All Compact Florescent Light-bulbs (CFL) and LED lamps contain them, and often the power supply fails before the light emitter.
Driving CFLs
LED Driver

Lightbulbs became electronics when they started to include switched mode power supplies.
All Compact Florescent Light-bulbs (CFL) and LED lamps contain them, and often the power supply fails before the light emitter.
Driving CFLs
LED Driver

Actually, if you look at this device, you'll see that gate-arrays aren't in the same class with your father's Oldsmobile any longer. We need them to be denser than the ones at that link, but the potential is there.

...but it uses an alternative proof-of-work algorithm called Equihash that supporters believe is impervious to being sped up with custom hardware

I laughed and laughed and laughed.

From the paper (PDF):

a reference implementation of a proof-of-work requiring 700 MB of RAM runs in 15 seconds on a 2.1 GHz CPU, increases the computations by the factor of 1000 if memory is halved, and presents a proof of just 120 bytes long.

Hmm... Needs less than 1 gigabyte. For external chips, that costs, let me see, $16 in modest quantities. Want it cheaper? Here is a magazine article from 4 years ago about people embedding memory in ASIC dies. In a modern chip process, 700 MB fits into what, a 5mm square?

They wave their hand over it in the paper, so it isn't like they ignored the cheapness-of-memory problem entirely. What if we assume that they are right and they have indeed found a problem with a critical dependency on memory throughput. Is there an obvious solution to that problem? What is the fastest memory in the world? (What word is entirely missing from their paper?) SRAM. In-die SRAM can be absurdly fast, like full core speed for arbitrary values of "core speed" and no wait cycles. It makes no sense to load a CPU up with piles of the stuff because caching has diminishing returns. But, what if your goal was to use ~6 billion cells of SRAM not as a cache, but as your main memory... How much faster would that ASIC be than a general purpose CPU?

Bottom line, if you imagine that you have a computation problem that can't be solved by building a single-purpose chip, you are almost certainly wrong. Either you don't understand your problem, or you don't understand the array of solutions available to solvers.

So, they hooked a solar cell to a thermoelectric cooling cell. Its the same tech as those USB refrigerators. You can build a basic a basic unit for $30-40.

I've thought about doing a similar thing to this to defeat ALPRs but it seems that most of these methods don't take it far enough. most of them are trying to get some bleed into the surrounding area of the sensor. A few watts of power draw is nothing which is what most of these attempts do, I'm thinking like 100W power draw for each license plate. I'm looking for this effect but in the IR. So instead of trying to create lens flare I want to massively underexpose the image. In this case I may also get some massive lens flare as well but that isn't what I would have been shooting for. Also using some LEDs like these would be good as they are far enough into the IR that they don't have the red glow that others do.

Any other SoC provider is free to offer a comparable cheap option. Let me know when that happens, OK?

TI does

Samsung does

There are a Whole slew of them

apowercap just came to mind. if you want small, high voltage caps, try tdk.

http://www.digikey.com/en/prod...

also, again, glass vial enclosure full of sulfur hexafluride. it does not conduct electricity, a least not in the ranges being discussed.
https://en.wikipedia.org/wiki/...

That that is an absolute lie, total BS. A bulky 3.5mm jack is 0.5cc. Assuming the previous life of their 1715mA/3.7V battery was 14 hours, they are now adding another 245mA/3.7V worth of battery to it, to get those 2 additional hours. That translates to an energy density of around 6.4 MJ/L, about three times what the best LiPo batteries can give. Not a chance.

If it lasts longer, it's not from a bigger battery, it's from more efficient components elsewhere. They're feeding you a line and you're swallowing it.

Here you go. It's in there somewhere. Good luck.

I looked for this on Digikeyand didn't find it, but they do have a bunch of SolidRun boards, including the ClearFog Pro which has 6+1 GbE and an SFP, plus 2 mPCIe slots and 1 mSATA. At less than $175, that's a steal. I've been waiting for something like this that could replace the discrete routers, ethernet switches, and Arduino programming slaves at my transmitter sites, which only have a few hosts each. There have been other multi-nic SBC's, but I have yet to see one this cheap.

I looked for this on Digikeyand didn't find it, but they do have a bunch of SolidRun boards, including the ClearFog Pro which has 6+1 GbE and an SFP, plus 2 mPCIe slots and 1 mSATA. At less than $175, that's a steal. I've been waiting for something like this that could replace the discrete routers, ethernet switches, and Arduino programming slaves at my transmitter sites, which only have a few hosts each. There have been other multi-nic SBC's, but I have yet to see one this cheap.

Toshiba FLASH 32GB (4GBx8), MMC interface - $5.38 in ones, $3.42 in 5k quantities. I spent about 45 seconds looking. This is not the cheapest available.

http://www.digikey.com/product...

By the way, 16GB (2GBx8) MMC is essentially the same price.

Might I suggest lots of high output IR LEDs in a hat maybe. To do a dazzle effect is hard but screwing up the exposure is easier if you can have an abnormally bright area. I have wanted to do somehting like that to automated license plate readers (ALPR) and some people have played around with that idea but they were trying to dazzle the reader where I want a larger area of high output LEDs to throw off the exposure, like a couple hundred watts of IR LED light.

Digikey has them with some minor volume discounts for 10 units.

Ebay, banggood or alibaba have them dirt cheap. A got a couple boards this way, and they're perfectly fine and well made. In particular on alibaba there are a bunch of $1/unit suppliers, though that's the FOB price and then you're on your own.

Or you just buy a ST Nucleo for around $10. It's a whole family of STM32 ARM Cortex MCUs available on the same board with Arduino Uno R3 compatible headers and every IO pin exposed.

Haha, this is the biggest "Linux!!!!" flamebait article... Anyway, yeah, it seems fairly obvious. The trend for years has been a move back to server-side processing and services that are very heavy on collected user data.

The thing is, it's not just Evul Micro$oft spying on you with a telescope - the spying comes wrapped in services that people actually want. For many people it's a tradeoff - they know they're giving their data away, but they're willing to give up a bit of privacy for the convenience that comes with the service. I find that the average person doesn't usually have the same reverence for privacy as the average Slashdotter.

It's already been said, and it'll be said a hundred more times before we let this article go, but yes, Linux and FOSS in general are the answer. We've been going back and forth about the Year of the Linux Desktop, but really, this is where FOSS shines: as a relatively minority choice for enthusiasts. Let people make their privacy tradeoff choice in peace, it's a perfectly valid choice to make if one most of us (myself included) find highly distasteful, and the rest of us can work on and use FOSS to our heart's content.

Coming from the hardware side too, as more of an EE guy than a programmer: OSHW is getting more and more possible. Powerful hardware that is amenable to use in open designs is becoming more available every year. I can jump over to DigiKey and buy an ARM chip that is capable of running Linux and has more computing power than some of my first desktop computers for $20. The chip designs themselves tend not to be open, but they do tend to be quite well documented - the high end is almost always closed and subject to NDA, but there is little pressure to move that line backwards, and as the high end moves forward, the devices available to the OSHW developer get better and better.

I don't think this is the end of computing privacy, I think this is just the logical conclusion of computers (read: the computers in your pocket!) becoming popular, and starting to work the way Average Joe expects them to. Enthusiasts will always be here, and I think this is the start of a new era for them.

Yes, but FRAM is not very high density. I use one (a small SPI FRAM, 4kB) to preserve state in case of power loss in things I design and it's fantastic for this purpose: when the power off interrupt comes from the supply monitoring chip, I still have a few ms to save all the data, which is more than I need. With Flash it would be impossible, and I can't afford to pre-erase just in case: this would kill endurance. Even EEPROM would be messy. FRAM (I might also try MRAM in the future) allow me to blast the data at a rate only limited by the SPI clock.
Example FRAM chip: FM25L04B
Example MRAM chip:MR25H256CDC
That said, the 1000 times faster claim may be for writes, in this case the write times of the new memory are counted in microseconds, which is not fast. FRAM and MRAM have similar read and write access times (actually FRAM read is destructive, but it's not a concern when endurance is 10^13 to 10^15 cycles, MRAM is not destructive AFAICT) but current capacities are not for mass storage.
I personnally suspect that one of the most important aspect of this memory is that it can be built as a 3D struture and does not even need a single transistor per cell, this is what will allow a reasonable cost even if individual cells are relatively large.

Yes, but FRAM is not very high density. I use one (a small SPI FRAM, 4kB) to preserve state in case of power loss in things I design and it's fantastic for this purpose: when the power off interrupt comes from the supply monitoring chip, I still have a few ms to save all the data, which is more than I need. With Flash it would be impossible, and I can't afford to pre-erase just in case: this would kill endurance. Even EEPROM would be messy. FRAM (I might also try MRAM in the future) allow me to blast the data at a rate only limited by the SPI clock.
Example FRAM chip: FM25L04B
Example MRAM chip:MR25H256CDC
That said, the 1000 times faster claim may be for writes, in this case the write times of the new memory are counted in microseconds, which is not fast. FRAM and MRAM have similar read and write access times (actually FRAM read is destructive, but it's not a concern when endurance is 10^13 to 10^15 cycles, MRAM is not destructive AFAICT) but current capacities are not for mass storage.
I personnally suspect that one of the most important aspect of this memory is that it can be built as a 3D struture and does not even need a single transistor per cell, this is what will allow a reasonable cost even if individual cells are relatively large.

Are you sure? http://www.digikey.com/product...

You work in the PCB industry, don't you? One of you pop up in every discussion about PCB pricing. "Oh, those PCBs are crap, just look at these PHP errors as evidence. Plus I can't find basic information that's easy-to-find because I can't be bothered to look at more than two pages on the web site. Also, the only reason you care about costs anyway is because you're a dumbass spending more money to make things than it would cost to buy them and you're blaming that on the cost of the PCBs and not your own stupidity. If you were making anything out of necessity you'd be willing to pay the higher prices."

My favorite was when someone told me "well, if you're price conscious, you're free to shop around." Yes, because I wouldn't be price conscious when faced with choices like receiving 12 PCBs for $14 in 3 weeks (including shipping time) from China vs. a U.S. manufacturer offering 100 PCBs for $317 in 4 weeks + shipping time, and much worse deals if I don't actually want a hundred boards. I'm willing to pay more for U.S. products, but unfortunately I'm not wealthy and so I have limits.

As for your concern over wasting money, the only PCB I've had manufacturered is my own version of this overpriced thing, and the final cost, components and all, resulted in a savings of $213.60 for the 12 boards I made vs. buying those boards from Digikey. ...not that I actually saved $213.60. I'm not wealthy enough to afford that. What I actually gained was having 12 copies of the board at all, rather than only having two copies of the one from Digikey and having to constantly swap those two boards between electronics projects and be unable to use more than two such projects at a time.

Not sure if you include all things 'micro' in your general derision, but MEMS devices have revolutionized some areas of electronics, particularly sensors. I bet you've got a MEMS-based accelerometer in your phone right now, and possibly a MEMS-based programmable oscillator.

If you have sleep apnea or asthma, your therapy device might contain a MEMS-based flow sensor.

Not sure if you include all things 'micro' in your general derision, but MEMS devices have revolutionized some areas of electronics, particularly sensors. I bet you've got a MEMS-based accelerometer in your phone right now, and possibly a MEMS-based programmable oscillator.

If you have sleep apnea or asthma, your therapy device might contain a MEMS-based flow sensor.

Not sure if you include all things 'micro' in your general derision, but MEMS devices have revolutionized some areas of electronics, particularly sensors. I bet you've got a MEMS-based accelerometer in your phone right now, and possibly a MEMS-based programmable oscillator.

If you have sleep apnea or asthma, your therapy device might contain a MEMS-based flow sensor.

In bulk it is easy to purchase microcontroller with 32MB of ram for $7 (though the closest my lazy digikey search turned up was $11.50 for 24M).

The cost of the microcontroller has a negligible cost on the nest retail price, and darn near every other embedded system sold to consumers.

No, ARM isn't more expensive. Try, for example, the ST Microelectronics STM32F030R8T6. That's a Cortex-M0 ARM, 48MHz, 64K flash, 8K RAM, 55 I/O pins. $2.22 in quantity one. Reference: http://www.digikey.com/product... That's just one part I happen to be familiar with; there may be even cheaper ARM alternatives out there.

Quantity one price of an ATMega328? $3.25. That's the surface mount version; the DIP is $3.38. Reference: http://www.digikey.com/product...

It's true that if you stay with Atmel, ARM will be more expensive. The ATSAMD21G18 that is used in the upcoming Arduino Zero Pro is $6.17 in quantity one. Reference: http://www.digikey.com/product... To be fair, that is a newer design using the somewhat more powerful Cortex-M0+ core.

No, ARM isn't more expensive. Try, for example, the ST Microelectronics STM32F030R8T6. That's a Cortex-M0 ARM, 48MHz, 64K flash, 8K RAM, 55 I/O pins. $2.22 in quantity one. Reference: http://www.digikey.com/product... That's just one part I happen to be familiar with; there may be even cheaper ARM alternatives out there.

Quantity one price of an ATMega328? $3.25. That's the surface mount version; the DIP is $3.38. Reference: http://www.digikey.com/product...

It's true that if you stay with Atmel, ARM will be more expensive. The ATSAMD21G18 that is used in the upcoming Arduino Zero Pro is $6.17 in quantity one. Reference: http://www.digikey.com/product... To be fair, that is a newer design using the somewhat more powerful Cortex-M0+ core.

No, ARM isn't more expensive. Try, for example, the ST Microelectronics STM32F030R8T6. That's a Cortex-M0 ARM, 48MHz, 64K flash, 8K RAM, 55 I/O pins. $2.22 in quantity one. Reference: http://www.digikey.com/product... That's just one part I happen to be familiar with; there may be even cheaper ARM alternatives out there.

Quantity one price of an ATMega328? $3.25. That's the surface mount version; the DIP is $3.38. Reference: http://www.digikey.com/product...

It's true that if you stay with Atmel, ARM will be more expensive. The ATSAMD21G18 that is used in the upcoming Arduino Zero Pro is $6.17 in quantity one. Reference: http://www.digikey.com/product... To be fair, that is a newer design using the somewhat more powerful Cortex-M0+ core.

All plates must be (1) securely fastened so as to prevent them from swinging, (2) displayed horizontally with the identifying numbers and letters facing outward from the vehicle, and (3) mounted in the upright position. The person driving the motor vehicle shall keep the plate legible and unobstructed and free from grease, dust, or other blurring material so that the lettering is plainly visible at all times. It is unlawful to cover any assigned letters and numbers or the name of the state of origin of a license plate with any material whatever, including any clear or colorless material that affects the plate's visibility or reflectivity.

So this makes it illegal to use the most effective method of stopping these things. Also it makes almost all of those vanity license plate holders illegal so consider that as it just gives them an excuse to pull you over. The other popular method of attempting to defeat ALPRs seems to be to use a few high output IR LEDs in attempt to dazzle the sensor by flooding surrounding pixels but in those don't seem to work all that well. Others have tried to flash IR LEDs but this may also be illegal in your state as it is in mine as there are usually laws governing flashing lights in motion.

My take would be instead to pump out enough IR over a large enough are to mess with the camera's exposure. Bye enough IR over a large enough of an area I am thinking of a few hundred watts over an area slightly larger than the license plate immediately surrounding it, but not covering the plate at all. From what I can tell this hasn't been attempted, but now we are talking real bright as the sun levels of power and that should screw up the metering on the camera. For LEDs I am thinking these as they appear to be about as high efficiency as I can find that put out a lot of IR.

Err - no.
Look at the URL, there is a clue why you're spectacularly wrong.

Current LEDs (blue ones, which white is based on) exceed 50% quantum efficiency.

http://www.digikey.com/product... - for example - does 48% electricity to light.

Slashdot replies are less likely to be sponsored.

That should be modded Funny, or Sad. But in truth Slashdot opinions are nearly as likely to be "astroturf" as legitimate compared to anywhere else.

Personally I do like mechanical keyboards, and for a non-backlit model the Cherry mechanical keyboard (not a 3rd-party keyboard using Cherry switches) I own and can recommend the G80-3000 (USB 104-keys US keymap), available Digikey and other (industrial) electronic suppliers globally. Likely just not your local / mail-order computer shop.

Otherwise for mechanical keyboards it is more a matter of selecting which key switch characteristics (resistance, push-length, noise, etc) than particular brands.

Of course any good keyboard discussion requires mentioning the IBM Model-M successors from UniComp with distinctive yet potentially annoyingly loud, buckling spring switches.

One of these or one of these if you want to do direct SMS...

All plates must be (1) securely fastened so as to prevent them from swinging, (2) displayed horizontally with the identifying numbers and letters facing outward from the vehicle, and (3) mounted in the upright position. The person driving the motor vehicle shall keep the plate legible and unobstructed and free from grease, dust, or other blurring material so that the lettering is plainly visible at all times. It is unlawful to cover any assigned letters and numbers or the name of the state of origin of a license plate with any material whatever, including any clear or colorless material that affects the plate's visibility or reflectivity.

So in my state any cover one were to put over their plate, including those license plate frames dealers put on, would be illegal. At the same time if one were to rig up a system where one was dumping out massive amounts of IR around the plate in an effort to flood the image that would be legal. Also for the record there are no laws covering the IR emission from vehicles in Minnesota as all light emissions regulations deal with specific colors (white, blue, amber, and red) light, or with flashing lights.

I have done some digging into seeing if flooding ALPRs with IR is possible and while some people seem to say it isn't it seems like their efforts have been fairly half assed. They only put out a few watts of power instead of going for a few hundred watts of power. I want to build a frame that covers no part of my plate but will draw 20 amps at 12V and dump it into a large array of these IR LEDs. From the pictures I have seen where people photograph a 100 equivalent watt bulb showing that even that doesn't flood the image they just didn't take it far enough so ~200W going into some IR LEDs would be about as bright as a 1000 watt bulb which now is starting to get up into the range of back lit by the sun range type of power which will mess with the picture.

FYI, Digi-Key will ship your order for free if you mail them a check or money order.

* When a check or money order accompanies your order, Digi-Key pays all shipping and insurance (our choice for method of shipping) to all addresses in the U.S. and Canada.

Section II.6

If you want to control a few motors and lights with network connectivity, get some ESP8266 modules - those are WiFi modules with a user-programmable 80MHz 32-bit CPU that you can buy for $5. Throw in a Cortex-M0 as a slave device to control your I/O (which can be as cheap as $1 in single quantities - yes, you can get a 32-bit CPU for $1 these days). That is what 2015 state-of-the-art silicon gets you to fit the task. A Raspberry Pi with a WiFi dongle is an order of magnitude more expensive and overpowered (and yet underpowered relative to what it claims to be, which is a Linux platform).

I have thought of building an LED license plate frame to mess with the cameras and others have tinkered with the idea some. The results at best could be considered hit or miss but that doesn't mean it couldn't be improved upon since most I have seen only output a few watts of power. I have been trying to figure out if I could build one with a power draw of 100-200W using some high output IR LEDs (the new license plates Minnesota uses are designed to be highly viable in the IR spectrum). Having a frame that isn't covering the license plate at all is perfectly legal in Minnesota but other things are not Minnesota statute 169.79 Subd. 7. Also the existing laws on vehicle illumination would also not prohibit this.

By frame I mean many concentric rings of LEDs packed tightly around the license plate so that you have 100 watts of IR LEDs shining around each plate. One of these days I will get some time to do it, and may also look into illuminating the front and rear windshield with IR LEDs as well in a similar fashion to further flood the image with IR.

Look into using a voltage regulator like these. I power my two RPis using them and since they have such a large input voltage range I can use lots of things to power them. I frequently use an old car battery to power one as a base station and an old laptop battery pack to power a mobile roving one (I do this when mapping trails). These regulator a fairly efficient and seem to be dead on accurate in their voltage.

You must not be looking in the correct places. The link below is a product page with datasheet, specs, etc.

DDR3 Module

Near-instant charging.

Irrelevant. You're still limited by supply rates and feed wire heating. Top end li-ion cells can charge in a matter of minutes on the small scale. In practice it's supply rate and cooling that limits you.

Much higher discharge rates

Irrelevant. What, you think cars have multi-megawatt inverters and motors? And again, top-end li-ions can have couple minut discharges.

and that without developing significant heat, because their series resistance is negligible

Slow charge and discharge of li-ions (normal usage) is usually over 99% efficiency. Fast charge is usually 94-97%. Fast discharge is irrelevant because the rest of the car can't handle using multiple megawatts at once (what, you think the car's going to get 0-100 times measured in milliseconds?) Older supercaps are less efficient than older li-ions due to an increase in the internal resistance (more in this in a minute)

Enormously more charge/discharge cycles than anything in battery tech

10 years-ish isn't good enough for you? Fine, reduce the depth of discharge to get 15-20 years. You'll still be an order of magnitude higher energy density than ultracaps.

you could will ultracaps used in a vehicle context to your children

You've been way overstating (and repeating a common mythology) about ultracaps. They don't actually last that long. Here's an info sheet from an ultracap manufacturer. Scroll down to "Life Expectancy".

The life expectancy of supercapacitors is identical to aluminum electrolytic capacitors ... Supercapacitors operated at room temperature can have life expectancies of several years

It's pure nonsense that they last forever. Some are rated for longer. For example Ioxus rates theirs at 10 years. But 10 years is pretty common for higher end EV battery packs, too.

This concept that ultracapacitors are something that you can "gift to your children" is just bull. They degrade, too. Following an exponential degradation curve dominated by increasing internal resistance.. So please stop with this nonsense.

Much wider range of usable performance over temperature; much colder, much hotter.

As per the above, operating out of the ideal temperature range cuts your ultracap lifespan. Commercial ultracaps aren't generally rated for wider temperature operating ranges than high-end li-ions, as you'll see from the various linked caps in this post, which are just a random sampling (for example, the Cooper Bussmann aerogel caps are only rated down to -25C, which is not impressive at all). And there's only a rather small range that's necessary for human-operated vehicles on the surface of the Earth. The ambient temperature outside isn't going to reach cold enough to liquify oxygen or melt zinc.

Much less need for recycling

Show me a single type of ultracap which can be recycled at all.

They can't be overcharged at their rated voltage

Overheating of the supercapacitor can occur from continuous overcurrent or overvoltage charging. Overheating can lead to increased ESR, gas generation, decreased lifetime, leakage, venting or rupture. .... General Safety Considerations: Supercapacitors may vent or rupture if overcharged, reverse charged, incinerated or heated above 150C

The

Heck, I don't even know who would need solder paste without actually using a laser-cut stencil and, you know, actually printing the paste like it was meant to be - in quantity? What's the point? You don't even need or want a fine-tipped soldering iron. For reflowing anything with leads, you in fact want a nice 3mm-5mm wide, short tip with good thermal conductivity. You don't need solder paste, you do need a flux pen .

I think this is correct. Sparkfun yellow looks too close to Fluke yellow to be a coincidence. Other manufacturers use 'safety colors' like this. Or this. Both of which seem to keep their respective manufacturers out of trademark trouble.

Retail electronics parts stores are dead. Even in Silicon Valley, we barely have any left. Digi-Key used to have a minimum order of $25. But they dropped that a few years ago. You can order one resistor from them and it will ship the same day by first class mail, in a small padded envelope. This pretty much solved the parts problem for people who know what they want.

The Digi-Key site can be overwhelming to hobbyists. Want a 100 ohm, 1/4W resistor for through-hole mounting? Radio Shack has one type. Digi-Key has 225 different types. That's part of what keeps Radio Shack and Jameco in business. If Digi-Key or Mouser ever sets up a hobbyist-friendly front end site to their inventory, the last need for the little guys will disappear.

The big question is, in order to ensure their survival, would Radio Shack be better off continuing to phase out their brick and mortar presence while making substantial efforts to expand as an exclusively online retailer?"

Not likely. They have no particular advantage in the online space aside from a recognizable (if tarnished) brand name. What they really should have done was to expand their catalog sales back in the day and become a distributor like Digikey or Mouser. I suppose they still could though they are behind the curve. They've gotten into cell phones but no one really thinks Radio Shack when they think cell phones. They sell batteries but there now are specialty battery stores that usually have a better selection and better prices. They don't have the scale or the expertise to compete with Amazon online and they are too unfocused to have profitable retail space. I can't really think of anything where Radio Shack would be my preferred shopping destination.

Radio shack has been trying to be all things to all people and when you do that you don't serve any of them well. They have expensive real estate, small square footage, small selections of products, high prices and unclear strategy. Their advantages are that they are fairly well known and have a lot of storefronts. That's a pretty thin advantage these days. I'm thinking Radio Shack might be a pretty good stock to think about short selling.

It's an ATTiny ($1.08 in quantities of 1) with V-USB (software only USB), most likely.

You could easily drive a strip, with animations and all, using a a cheap shift register or a dedicated led array driver.

Heck this isn't even new:

http://www.yageo.com/exep/pages/download/literatures/an08436.pdf

and to give a slightly different view:

http://www.digikey.com/Web%20Export/Supplier%20Content/TDK_445/PDF/tdk_high_cap_replacement.pdf?redirected=1

that is just as old.

Ridiculously there is something new as well:

http://www.element14.com/community/servlet/JiveServlet/downloadBody/55139-102-1-275420/Panasonic%20Polymer%20Capacitor%20Technologies%20-%20Replacement%20of%20Tantalum%20and%20MLCC.pdf

You can bump the switch with your elbow when your arms are full

This light bulb has to be always ON at the switch. Perhaps you can cycle it off and on, and then it will default to "on" after application of power. But that's inconvenient and counter-intuitive.

you can have the lights turn on when your phone is in range after you come home

Perhaps that will work if you are a King and live in a palace, with rooms that are far apart. Modern homes have rooms packed close to each other, and drywall does not block RF. You will end up activating lights in all adjacent rooms, regardless of the need.

automatically turn on when your garage opens and it's dark at the end of the day

My garage opener already has two light bulbs (LED) that activate for 5 minutes after the door moves. That's plenty of time to park the car, grab your things and get into the house.

you can turn them on from a button in your car

Once you get into this level of functionality you want to have an established home automation system, like Z-Wave or Insteon. They already have portable and stationary controllers, and a bunch of devices that can be controlled. Is this company on Kickstarter is trying to compete with Z-Wave? They'd become yet another proprietary solution that is a thing in itself. The huge advantage of Z-Wave is that it is a standard, and everyone can make Z-Wave devices that will work with everyone else's Z-Wave devices. For that purpose Z-Wave has published a set of classes, where the programming instructions for every class (sensor, light switch, etc.) are explained in detail. Technology-wise, Z-Wave is also a mesh, it runs in the 900 MHz ISM band, and it does not interfere with your 802.11 networks. Insteon is a mixed (power line and wireless) technology. Both are pretty inexpensive.

But I get it - smart phones are complicated and those damn kids use them too much already so no one should have more technology - especially not nerds on a tech news site.

You want more than one controller at your home. Do you want to pay for several smartphones that are just laying around the house so that visitors can use them? Do you *always* have your smartphone on you? (I don't even own one, and I have no plans to get such a thing.) What if you forget your smartphone at work, or lose it, or break it?

Intel might be the first to do it on a CPU die, but they're not the first to do on-silicon inductors by any stretch. Switching regulators with inductors on silicon have been commercially available for several years now. The R-78 and MIC33030, for example, are drop-in replacements for linear regulators, with all components on die.

The real question in my mind is why anyone still uses linear regulators for anything, but I digress.