We use EC2 as the back-end for Netalyzr (our free, applet-based network testing and debugging service), and right now are in the middle of a minor flashcrowd with our big updated release. No recent glitches we've noticed, with long running small instances.
Each device needs a unique serial number, something to identify it. But at the same time, they didn't want to customize the firmware for each device to include a serial number.
So instead, some brilliant programmer observed that the embedded processor can get the MAC address from the NIC and use that as a serial number for accessing the web page.
This is an old and useful trick, but the only problem is although it gives you a unique serial number per device, it gives you a predictable serial number per device and because of the nature of the back-end service, they didn't just need a UNIQUE serial number, but also an UNPREDICTABLE serial number. Ooops.
No, the Israeli worry was that she was a terrorist DUPE, and was acting like one: she fit that profile to a T.
EG, this woman is acting very similar to the case in 1996 where a Jordanian tricked his pregnant Irish girlfriend into trying to carry a bomb onto an El Al flight.
That is a VERY old link, back when an iPod Touch was $400, not $200.
The bill of materials has not gone down by much in price: its still pretty close to $150 for materials, assembly, test, and shipping from China, especially when you consider the cost of the touchscreen.
And selling outside of the direct to consumer channel, Apple has to be selling it to the retailers for not that much more. The profit on the iPod touch at $200 is Not Much, probably the lowest of any major device Apple has ever sold.
In fact, it wouldn't suprise me if the $200, 8 GB touch has a close to zero profit to apple when sold through a non-Apple retailer... That is, until the buyer start loading apps on it.
Except most of the cost is the screen. This isn't "miniaturization is expensive" but actually "small is cheap": a 12" capacitive touchscreen is a big BIG money item.
The CrunchPad model only made sense at the low price: something inexpensive and universal.
But the low price never made sense. Apple doesn't make much money on the iPod touch, and they have all the huge economies of scale, and its still costs $200 for the 8 GB model. Add in a MUCH larger screen and bigger battery and of course the price will balloon.
LCDs generally use a lot less power than plasma TVs.
LCDs with LED backlights are even better... Those TVs already meet the 2013 california specifications.
EG, the Vizio 55" LCD tv with LED backlights draws only 150W average. So significantly bigger LCD backlit TV (20% larger area) draws only 20% of the power of a plasma TV.
The standards are not only necessary (its a suprisingly large fraction of the household power consumption in CA), but imminently doable.
Roughly 25% of the TVs on the market ALREADY meet the 2013 specification, with 50% meeting the 2011 specification.
The key is "LCD with LED backlight". Such TVs easily meet the spec and are of good quality.
LCD's with conventional backlights needs to change the backlight technology, but they are doing this anyway: LED backlights are better for longevity as well as power consumption.
Who this hurts is those who have bet on Plasma technology, as plasma can effectively not meet these requirements, but plasma is dying anyway, as LCD screens keep getting bigger and faster reacting while being cheaper than plasma TVs.
All the FPGA vendors have their own embedded CPU cores, such as the Xilinx Microblaze and the Altera NIOS II which are very small FPGA-embeddable CPU cores.
One of the reasons why BitTorrent didn't suffer the legal fate of Napster, Kazaa, etc is that BitTorrent only handles data transfer, not search, and has significant noninfringing uses.
Having trackerless torrents however doesn't help the noninfringing uses, only infringing uses. (If its non-infringing, just host a tracker damnit!), thus trackerless client features start to get very dangerous from a legal perspective for the developers.
Lets do a little math. Good video over the net is 2 Mbps for Netflix. At that rate, this is ~9 hours of video a DAY before you get to the 250 GB cap. Do you watch 9 hours of video a DAY over netflix's service?
Time/Warner's previous attempts to do a 50 GB cap? Thats anticompetitive.
But comcast's is sooo high that you basically have to be a massive Warez trader or doing something very stupid (offsite backup better handled by Sneakernet) to get to.
And its not throttling, its a fairness mechanism: It means that light users won't get outcompeted by heavy users, but heavy users shouldn't get starved out unless things are really REALLY bad.
Slashdot Mirror
We use EC2 as the back-end for Netalyzr (our free, applet-based network testing and debugging service), and right now are in the middle of a minor flashcrowd with our big updated release. No recent glitches we've noticed, with long running small instances.
As has been mentioned by others, WiFi sensitivity should easily count for paranormal under the James Randi Educational Foundation's $1M paranormal prize. http://www.randi.org/site/index.php/1m-challenge.html
He killed me with a sword. How weird is that?
Its sloppy to do, but here's why they did it....
Each device needs a unique serial number, something to identify it. But at the same time, they didn't want to customize the firmware for each device to include a serial number.
So instead, some brilliant programmer observed that the embedded processor can get the MAC address from the NIC and use that as a serial number for accessing the web page.
This is an old and useful trick, but the only problem is although it gives you a unique serial number per device, it gives you a predictable serial number per device and because of the nature of the back-end service, they didn't just need a UNIQUE serial number, but also an UNPREDICTABLE serial number. Ooops.
You really should sue for $47 Bazillion dollars...
It has a much better ring to it than $500K.
Specifically, why it ends up so bad on dropping VOICE calls. Yes, data is a huge hog. But the bulk data can be prioritized much lower than voice.
I could see in Manhattan or San Francisco that the DATA service would suck: too many users, etc. But why should the voice channel suck so badly too?
No, the Israeli worry was that she was a terrorist DUPE, and was acting like one: she fit that profile to a T.
EG, this woman is acting very similar to the case in 1996 where a Jordanian tricked his pregnant Irish girlfriend into trying to carry a bomb onto an El Al flight.
Anne-Mary Murphy
Are the lawyers. The Crunchpad/JooJoo was doomed from the start: too expensive given the current technology...
So all that will happen out of this is entertaining lawsuits where the laywers make their money and everyone else just laughs.
That is a VERY old link, back when an iPod Touch was $400, not $200.
The bill of materials has not gone down by much in price: its still pretty close to $150 for materials, assembly, test, and shipping from China, especially when you consider the cost of the touchscreen.
And selling outside of the direct to consumer channel, Apple has to be selling it to the retailers for not that much more. The profit on the iPod touch at $200 is Not Much, probably the lowest of any major device Apple has ever sold.
In fact, it wouldn't suprise me if the $200, 8 GB touch has a close to zero profit to apple when sold through a non-Apple retailer... That is, until the buyer start loading apps on it.
Except most of the cost is the screen. This isn't "miniaturization is expensive" but actually "small is cheap": a 12" capacitive touchscreen is a big BIG money item.
The same applies to batteries.
For a non-academic conference, it needs to be
a: A good enough topic to convince the boss to pay...
b: Cheap enough to convince the boss to pay...
c: In a nice enough location that you want to go...
So a $100/person conference in Hawaii sounds about right to me.
The CrunchPad model only made sense at the low price: something inexpensive and universal.
But the low price never made sense. Apple doesn't make much money on the iPod touch, and they have all the huge economies of scale, and its still costs $200 for the 8 GB model. Add in a MUCH larger screen and bigger battery and of course the price will balloon.
The Kindle for all its annoyances attempts to be as customer centric as possible.
But this, "publisher centric" model seems really l8me... EG, advertisements built into the fabric?
Mystery, unavailable devices?
An over-leveraged print empire driving it?
* (carbon, code, whatever) offsets are really the Papal indulgences of the 21st century.
So I don't think I'll look at the article until I actually need to program in Haskel....
Thats the mercury backlight in a conventional LCD. Those are being replaced by LED arrays.
LCDs generally use a lot less power than plasma TVs.
LCDs with LED backlights are even better... Those TVs already meet the 2013 california specifications.
EG, the Vizio 55" LCD tv with LED backlights draws only 150W average. So significantly bigger LCD backlit TV (20% larger area) draws only 20% of the power of a plasma TV.
The standards are not only necessary (its a suprisingly large fraction of the household power consumption in CA), but imminently doable.
Roughly 25% of the TVs on the market ALREADY meet the 2013 specification, with 50% meeting the 2011 specification.
The key is "LCD with LED backlight". Such TVs easily meet the spec and are of good quality.
LCD's with conventional backlights needs to change the backlight technology, but they are doing this anyway: LED backlights are better for longevity as well as power consumption.
Who this hurts is those who have bet on Plasma technology, as plasma can effectively not meet these requirements, but plasma is dying anyway, as LCD screens keep getting bigger and faster reacting while being cheaper than plasma TVs.
Don't worry about the encryption overhead, just make sure AES is the first cypher chosen. AES is VERY fast.
Set up one computer as a master repository and do everything over SVN+SSH.
Its what I do for the complete working set I have, passing between 3+ systems, is everything is through subversion over SSH to a backed-up system.
All the FPGA vendors have their own embedded CPU cores, such as the Xilinx Microblaze and the Altera NIOS II which are very small FPGA-embeddable CPU cores.
You also have free options that aren't tied to specific FPGAs like the LEON sparc-compatible processors.
One of the reasons why BitTorrent didn't suffer the legal fate of Napster, Kazaa, etc is that BitTorrent only handles data transfer, not search, and has significant noninfringing uses.
Having trackerless torrents however doesn't help the noninfringing uses, only infringing uses. (If its non-infringing, just host a tracker damnit!), thus trackerless client features start to get very dangerous from a legal perspective for the developers.
If you don't like your job, get a new one and then as you leave, snitch to the BSA for the bounty money.
Lets do a little math. Good video over the net is 2 Mbps for Netflix. At that rate, this is ~9 hours of video a DAY before you get to the 250 GB cap. Do you watch 9 hours of video a DAY over netflix's service?
Time/Warner's previous attempts to do a 50 GB cap? Thats anticompetitive.
But comcast's is sooo high that you basically have to be a massive Warez trader or doing something very stupid (offsite backup better handled by Sneakernet) to get to.
Comcast rolled this out nearly a year ago.
And its not throttling, its a fairness mechanism: It means that light users won't get outcompeted by heavy users, but heavy users shouldn't get starved out unless things are really REALLY bad.