The article you quote (including the section you quote) indicates that
1) Boeing does not think there is interference from cell phones; all emissions outside the cell phone's operating frequency were within airplane equipment emissions limits. Airplane communications and navigation frequencies are separate from cell phone operating frequencies.
"Generally within" emissions limits, not a blanket within. The fact that the sample 16 phones did not exceed emission limits is hardly any guarantee that other phones do not, nor that one of those sampled phones with replacement firmware would not. The fact that they are emitting at all on navigation and landing frequencies is a reason to keep an open mind regarding interference.
2) Anomalies observed in flight are NOT being reproduced in the lab, despite efforts to do so.
I thought I said that after the quotes, but when I reread I see that I left out the word "not". Sorry for the confusion.
In any case I hardly think that dismisses observed behavior such as "A passenger’s palmtop computer was reported to cause the airplane to initiate a shallow bank turn. One minute after turning the PED off, the airplane returned to "on course." When the unit was brought to the flight deck, the flight crew noticed a strong correlation by turning the unit back on and watching the anomaly return, then turning the unit off and watching the anomaly stop. "
To translate things into terms that the slashdot audience may have an easier time understanding: The failure to reproduce a software bug on the programmer's system is hardly evidence that the software is fine.
"Boeing conducted a laboratory and airplane test with 16 cell phones typical of those carried by passengers, to determine the emission characteristics of these intentionally transmitting PEDs. The laboratory results indicated that the phones not only produce emissions at the operating frequency, but also produce other emissions that fall within airplane communication/navigation frequency bands (automatic direction finder, high frequency, very high frequency [VHF] omni range/locator, and VHF communications and instrument landing system [ILS]). Emissions at the operating frequency were as high as 60 dB over the airplane equipment emission limits, but the other emissions were generally within airplane equipment emission limits. One concern about these other emissions from cell phones is that they may interfere with the operation of an airplane communication or navigation system if the levels are high enough."
"Operators of commercial airplanes have reported numerous cases of portable electronic devices affecting airplane systems during flight. These devices, including laptop and palmtop computers, audio players/recorders, electronic games, cell phones, compact-disc players, electronic toys, and laser pointers, have been suspected of causing such anomalous events as autopilot disconnects, erratic flight deck indications, airplanes turning off course, and uncommanded turns. Boeing has recommended that devices suspected of causing these anomalies be turned off during critical stages of flight (takeoff and landing). The company also recommends prohibiting the use of devices that intentionally transmit electromagnetic signals, such as cell phones, during all phases of flight."
The problem seems to be that anomalies observed in flight are being reproduced in a lab.
Are allowed to use iPads in all phases of flight, but if I read a book on mine... It's trouble.
The pilots may have wifi only iPads, no cellular. The pilots can probably be trusted to put the device into airplane mode during takeoff and landing, or to turn them off. The iPads are replacing printed documentation for aircraft info, maps, etc; they may very well be off during takeoff and landing.
I flew a few days ago. We were only instructed to turn off devices during takeoff and landing, not the entire flight.
This is what I was thinking. However I think you're being a little naive about what it will take to get a BS doing part time coursework.
I worked about 30 hours a week during my BS CS while taking at least 3 classes per quarter. I was busy but it was not too bad. I could have done 40 hours at work with some effort. Admittedly I had some flexibility in my schedule so I could make up hours early, late or weekends. Without this flexibility at work it would have been impossible, I was very fortunate in this regard.
I am not suggesting part-time coursework. IIRC full-time was 12 units (3 classes). If I had made better use of my local junior college I could have done more general ed at night and had to shift hours around at work a bit less. And of course taking classes during the summer session would be needed. IIRC I had 7 years to complete the requirements for the BS.
Since I was working as a programmer I was also able to get work to count for some of my elective units. Some type of cooperative education class where all I needed to do was write a report at the end of the quarter regarding what I was doing at work and how it incorporated recognized fields of study or topics in CS.
Again, I am not saying it is easy, just plausible.
BS CS programs typically make no accommodations for students working full time. There may be the odd class at 6pm but that has more to do with scheduling professors than anything else.
If your work will let you juggle your hours around a little so that you can take a class during the day and make up that time early morning, evening or weekends then you may be able to pull it off. The trick is to take as many general ed classes as you can at night. Junior colleges are especially helpful in this regard, just make sure the classes are fully transferable, double-check with the 4-year school. And of course keep an eye out for the occasional 6pm CS class.
Its tricky but if work, the 4-year and the JC are in relatively close proximity and if work can be a little flexible with hours two or three days a week it is plausible.
Corporations aren't inherently evil;) but from the practical side, usually the tuition payback is spread out over a few years, or is done like a signing bonus where you pay it back if you leave within a certain period of time.
If you are leaving early and are supposed to pay back a tuition reimbursement ask the HR folks to waive the requirement to repay. They may agree to do so. Get their answer on paper, if email print it out with full headers shown.
It does not hurt to ask. I've been pleasantly surprised and received such a waiver.
... new black and white films come out all the time...
Yes, but black and white films are not degraded and faded versions of color films.
Perhaps a more appropriate analogy would be to convert a color film to grayscale and to add some static to simulate a small black and white TV and its integrated antenna.:-)
Seriously. Cell phones finally started getting decent cameras to take pictures with so now people run filters to make them look like faded polaroids?
Are there 8-track emulator apps yet that play a song from your digital library, fade out in the middle, play a loud mechanical thunk sound effect, and then fade the song back in?
When vendors don't publish drivers or specs how is that supposed to happen?
You sign the vendor's non-disclosure agreement and then they provide the specs. So in a way this is a "cost" of being open source, some won't buy into your model. Everything has a cost.
FWIW, as many others have commented, I've been buying Intel motherboards and NICs for over a decade and I've never had a problem.
And one step back from an obvious crash may be in the subtle errors region where CPU failures can't be easily distinguished from software bugs. For example the subtle error can simply be an erroneous answer, 2+2=5 sort of stuff
That's why you run utilities like SuperPi. You can check that erroneous answer against known values. Benchmarks like this are designed to be as stressful as possible on the CPU, so that there's good reason to expect that if you don't see errors in SuperPi you won't see them anywhere else.
Honestly, that is not the case. These subtle erroneous answers can be dependent upon only a certain sequence of instructions being executed immediately before or certain data patterns being operated upon immediately before and the sequences or patterns may be different for each individual CPU, and the instruction offering the erroneous answer may vary for each individual CPU as well.
These utilities can only help detect the errors towards the more serious end of the spectrum, not the more subtle.
The typical method used by hobbyists was: overclock step by step until it crashes. Then, back off one step. You are now at the "optimal" speed - i.e. the fastest and therefore best speed. Games crash? Must be software bugs!
And one step back from an obvious crash may be in the subtle errors region where CPU failures can't be easily distinguished from software bugs. For example the subtle error can simply be an erroneous answer, 2+2=5 sort of stuff. If that erroneous answer is part of the calculation of where to draw something on the screen the error may be of no consequence, one pixel off may be imperceptible. However if that erroneous answer is ultimately part of the calculation of an array index then being one index off may be an array overflow and result in a crash.
Re:I don't believe 1% of computers give wrong answ
on
Whose Bug Is This Anyway?
·
· Score: 3, Interesting
While at a large game company I wrote the code that collected CPU make and model, video make and model, amount of RAM, OS version, etc. Basically the type of info you see under minimum system requirements. The CPUID instruction can return a vendor string indicating who made the CPU. Intel CPUs return "GenuineIntel". On very very rare and often transient occasions the reported string had a misspelling, the misspellings generally indicated a single bit error. Whether an overclocked CPU generating subtle errors or bad RAM or a bad power supply or something else is responsible I can't say. All I really know is that outside of the CPU manufacturing facility things do go wrong in hardware. The article is consistent with various things I have seen.
The problem is that you can't really test an overclocked CPU. CPUs do not work perfectly or catastrophically fail (crash the software). There is a range of errors between these two extremes. Some of the overclocking induced errors are quite subtle, a simple incorrect answer. The problem is that these subtle errors are sometimes dependent upon a certain sequence of instructions or certain data patterns and the sequences and data, as well as the failing instruction and the clock speed where these errors manifest, can vary from one individual CPU to the next.
You can run all sorts of test utilities and you will probably only discover the CPUs with the more severe problems. You will probably not be able to tell the CPUs that have subtle failures from CPUs that are functioning properly.
The article suggests that the app does have a sign-up.
"Recall that Microsoft, TNW has learned, offered to remove all paid subscription options from its application, a compromise that was rejected. TNW suspects that the sticking point in that matter was the ability to sign up for SkyDrive through the application, a ‘no’ in Apple books."
I expect people will have Office on their computer and mobile device. So do the subscription purchases from the computer.
Exactly. I really don't see why Microsoft is making such a huge stink over this, if the in-app purchase of a subscription is the real issue for them.
Perhaps MS was thinking of offering the Mac version of Office through Apple's Mac App Store. That might trigger the same restrictions as the iOS App Store.
Apple's position is, and always has been, that subscriptions bought through in-app purchase pay Apple the 30% cut—because they're going through Apple's payment process and paying through their iTunes account—while subscriptions bought elsewhere, Apple doesn't care about.
I think all sales done by an app have to go through Apple's in-app purchase system. You can offer subscriptions and services elsewhere but the purchase has to be done outside of the app. I don't think the app can even send you to the web for such purchases to avoid in-app.
Now for MS Office this seems like a minor issue. I expect people will have Office on their computer and mobile device. So do the subscription purchases from the computer.
As much as I hate Microsoft, I'm behind them here. Perhaps they should just drop all their apps suddenly and promote the Android and Windows Phone versions?
I don't think too many users are going to be using MS Office only on their iOS devices, rather they will have MS Office on their computer and devices. So don't sell/manage the subscriptions from the App Store apps. Only sell/manage the subscriptions from the web or from computer based apps sold outside of the Apple App Store. Unlike iOS, the Apple App Store is not required under Mac OS X and developers may sell directly to users. If you are an established and well known brand and have the infrastructure to do your own digital sales to Windows and Mac OS X users then there is little need for the Apple App Store.
No, developers must release the source code to ANYONE who requests it, regardless of whether they received the binary or not. Read clause 3(b), the part where it says "any third party". The key word is "any".
No, the key words are "third party". A person who does not posses the binary is *not* a party to this contract. "Any" is merely a modifier on "third party" indicating that it does not matter how the third party received the binary, i.e. receiving from someone other than the developer.
What about the rights of the people who worked on it in the past, with the understanding that their "compensation" would be in the form of others donating their time to continue improvements?
What "rights" or "understandings" of the past developers? Their right, their understandings, are defined by the GPL that they chose to license their software under. The GPL allows for payment. The only thing the GPL disallows is denying access to the source code.
the only way google did it in kansas city was by paying off local officials to allow them to put their fiber on the poles at lower rates than everyone else's lines. not going to work everywhere. ISP's and others are already suing kansas city for allowing this
Giving local officials some sort of deal seems to be a well established practice that has worked in many locales. See the cell phone industry. A cellular tower is opposed until the provider offers to put equipment to support local police and fire communications up there. I suspect that the initial opposition is just a gambit to get such freebies in some locales. I'd be surprised if such practices have not already been ruled on by the courts.
Slashdot Mirror
The article you quote (including the section you quote) indicates that 1) Boeing does not think there is interference from cell phones; all emissions outside the cell phone's operating frequency were within airplane equipment emissions limits. Airplane communications and navigation frequencies are separate from cell phone operating frequencies.
"Generally within" emissions limits, not a blanket within. The fact that the sample 16 phones did not exceed emission limits is hardly any guarantee that other phones do not, nor that one of those sampled phones with replacement firmware would not. The fact that they are emitting at all on navigation and landing frequencies is a reason to keep an open mind regarding interference.
2) Anomalies observed in flight are NOT being reproduced in the lab, despite efforts to do so.
I thought I said that after the quotes, but when I reread I see that I left out the word "not". Sorry for the confusion.
In any case I hardly think that dismisses observed behavior such as "A passenger’s palmtop computer was reported to cause the airplane to initiate a shallow bank turn. One minute after turning the PED off, the airplane returned to "on course." When the unit was brought to the flight deck, the flight crew noticed a strong correlation by turning the unit back on and watching the anomaly return, then turning the unit off and watching the anomaly stop. "
To translate things into terms that the slashdot audience may have an easier time understanding: The failure to reproduce a software bug on the programmer's system is hardly evidence that the software is fine.
Boeing thinks there is interference:
"Boeing conducted a laboratory and airplane test with 16 cell phones typical of those carried by passengers, to determine the emission characteristics of these intentionally transmitting PEDs. The laboratory results indicated that the phones not only produce emissions at the operating frequency, but also produce other emissions that fall within airplane communication/navigation frequency bands (automatic direction finder, high frequency, very high frequency [VHF] omni range/locator, and VHF communications and instrument landing system [ILS]). Emissions at the operating frequency were as high as 60 dB over the airplane equipment emission limits, but the other emissions were generally within airplane equipment emission limits. One concern about these other emissions from cell phones is that they may interfere with the operation of an airplane communication or navigation system if the levels are high enough."
"Operators of commercial airplanes have reported numerous cases of portable electronic devices affecting airplane systems during flight. These devices, including laptop and palmtop computers, audio players/recorders, electronic games, cell phones, compact-disc players, electronic toys, and laser pointers, have been suspected of causing such anomalous events as autopilot disconnects, erratic flight deck indications, airplanes turning off course, and uncommanded turns. Boeing has recommended that devices suspected of causing these anomalies be turned off during critical stages of flight (takeoff and landing). The company also recommends prohibiting the use of devices that intentionally transmit electromagnetic signals, such as cell phones, during all phases of flight."
The problem seems to be that anomalies observed in flight are being reproduced in a lab.
http://www.boeing.com/commercial/aeromagazine/aero_10/interfere_textonly.html
Are allowed to use iPads in all phases of flight, but if I read a book on mine... It's trouble.
The pilots may have wifi only iPads, no cellular. The pilots can probably be trusted to put the device into airplane mode during takeoff and landing, or to turn them off. The iPads are replacing printed documentation for aircraft info, maps, etc; they may very well be off during takeoff and landing.
I flew a few days ago. We were only instructed to turn off devices during takeoff and landing, not the entire flight.
This is what I was thinking. However I think you're being a little naive about what it will take to get a BS doing part time coursework.
I worked about 30 hours a week during my BS CS while taking at least 3 classes per quarter. I was busy but it was not too bad. I could have done 40 hours at work with some effort. Admittedly I had some flexibility in my schedule so I could make up hours early, late or weekends. Without this flexibility at work it would have been impossible, I was very fortunate in this regard.
I am not suggesting part-time coursework. IIRC full-time was 12 units (3 classes). If I had made better use of my local junior college I could have done more general ed at night and had to shift hours around at work a bit less. And of course taking classes during the summer session would be needed. IIRC I had 7 years to complete the requirements for the BS.
Since I was working as a programmer I was also able to get work to count for some of my elective units. Some type of cooperative education class where all I needed to do was write a report at the end of the quarter regarding what I was doing at work and how it incorporated recognized fields of study or topics in CS.
Again, I am not saying it is easy, just plausible.
BS CS programs typically make no accommodations for students working full time. There may be the odd class at 6pm but that has more to do with scheduling professors than anything else.
If your work will let you juggle your hours around a little so that you can take a class during the day and make up that time early morning, evening or weekends then you may be able to pull it off. The trick is to take as many general ed classes as you can at night. Junior colleges are especially helpful in this regard, just make sure the classes are fully transferable, double-check with the 4-year school. And of course keep an eye out for the occasional 6pm CS class.
Its tricky but if work, the 4-year and the JC are in relatively close proximity and if work can be a little flexible with hours two or three days a week it is plausible.
Corporations aren't inherently evil ;) but from the practical side, usually the tuition payback is spread out over a few years, or is done like a signing bonus where you pay it back if you leave within a certain period of time.
If you are leaving early and are supposed to pay back a tuition reimbursement ask the HR folks to waive the requirement to repay. They may agree to do so. Get their answer on paper, if email print it out with full headers shown.
It does not hurt to ask. I've been pleasantly surprised and received such a waiver.
... new black and white films come out all the time ...
Yes, but black and white films are not degraded and faded versions of color films.
:-)
Perhaps a more appropriate analogy would be to convert a color film to grayscale and to add some static to simulate a small black and white TV and its integrated antenna.
Seriously. Cell phones finally started getting decent cameras to take pictures with so now people run filters to make them look like faded polaroids?
Are there 8-track emulator apps yet that play a song from your digital library, fade out in the middle, play a loud mechanical thunk sound effect, and then fade the song back in?
When vendors don't publish drivers or specs how is that supposed to happen?
You sign the vendor's non-disclosure agreement and then they provide the specs. So in a way this is a "cost" of being open source, some won't buy into your model. Everything has a cost.
FWIW, as many others have commented, I've been buying Intel motherboards and NICs for over a decade and I've never had a problem.
And one step back from an obvious crash may be in the subtle errors region where CPU failures can't be easily distinguished from software bugs. For example the subtle error can simply be an erroneous answer, 2+2=5 sort of stuff
That's why you run utilities like SuperPi. You can check that erroneous answer against known values. Benchmarks like this are designed to be as stressful as possible on the CPU, so that there's good reason to expect that if you don't see errors in SuperPi you won't see them anywhere else.
Honestly, that is not the case. These subtle erroneous answers can be dependent upon only a certain sequence of instructions being executed immediately before or certain data patterns being operated upon immediately before and the sequences or patterns may be different for each individual CPU, and the instruction offering the erroneous answer may vary for each individual CPU as well.
These utilities can only help detect the errors towards the more serious end of the spectrum, not the more subtle.
The typical method used by hobbyists was: overclock step by step until it crashes. Then, back off one step. You are now at the "optimal" speed - i.e. the fastest and therefore best speed. Games crash? Must be software bugs!
And one step back from an obvious crash may be in the subtle errors region where CPU failures can't be easily distinguished from software bugs. For example the subtle error can simply be an erroneous answer, 2+2=5 sort of stuff. If that erroneous answer is part of the calculation of where to draw something on the screen the error may be of no consequence, one pixel off may be imperceptible. However if that erroneous answer is ultimately part of the calculation of an array index then being one index off may be an array overflow and result in a crash.
While at a large game company I wrote the code that collected CPU make and model, video make and model, amount of RAM, OS version, etc. Basically the type of info you see under minimum system requirements. The CPUID instruction can return a vendor string indicating who made the CPU. Intel CPUs return "GenuineIntel". On very very rare and often transient occasions the reported string had a misspelling, the misspellings generally indicated a single bit error. Whether an overclocked CPU generating subtle errors or bad RAM or a bad power supply or something else is responsible I can't say. All I really know is that outside of the CPU manufacturing facility things do go wrong in hardware. The article is consistent with various things I have seen.
The problem is that you can't really test an overclocked CPU. CPUs do not work perfectly or catastrophically fail (crash the software). There is a range of errors between these two extremes. Some of the overclocking induced errors are quite subtle, a simple incorrect answer. The problem is that these subtle errors are sometimes dependent upon a certain sequence of instructions or certain data patterns and the sequences and data, as well as the failing instruction and the clock speed where these errors manifest, can vary from one individual CPU to the next.
You can run all sorts of test utilities and you will probably only discover the CPUs with the more severe problems. You will probably not be able to tell the CPUs that have subtle failures from CPUs that are functioning properly.
The article suggests that the app does have a sign-up.
"Recall that Microsoft, TNW has learned, offered to remove all paid subscription options from its application, a compromise that was rejected. TNW suspects that the sticking point in that matter was the ability to sign up for SkyDrive through the application, a ‘no’ in Apple books."
I expect people will have Office on their computer and mobile device. So do the subscription purchases from the computer.
Exactly. I really don't see why Microsoft is making such a huge stink over this, if the in-app purchase of a subscription is the real issue for them.
Perhaps MS was thinking of offering the Mac version of Office through Apple's Mac App Store. That might trigger the same restrictions as the iOS App Store.
Apple's position is, and always has been, that subscriptions bought through in-app purchase pay Apple the 30% cut—because they're going through Apple's payment process and paying through their iTunes account—while subscriptions bought elsewhere, Apple doesn't care about.
I think all sales done by an app have to go through Apple's in-app purchase system. You can offer subscriptions and services elsewhere but the purchase has to be done outside of the app. I don't think the app can even send you to the web for such purchases to avoid in-app.
Now for MS Office this seems like a minor issue. I expect people will have Office on their computer and mobile device. So do the subscription purchases from the computer.
As much as I hate Microsoft, I'm behind them here. Perhaps they should just drop all their apps suddenly and promote the Android and Windows Phone versions?
I don't think too many users are going to be using MS Office only on their iOS devices, rather they will have MS Office on their computer and devices. So don't sell/manage the subscriptions from the App Store apps. Only sell/manage the subscriptions from the web or from computer based apps sold outside of the Apple App Store. Unlike iOS, the Apple App Store is not required under Mac OS X and developers may sell directly to users. If you are an established and well known brand and have the infrastructure to do your own digital sales to Windows and Mac OS X users then there is little need for the Apple App Store.
No, developers must release the source code to ANYONE who requests it, regardless of whether they received the binary or not. Read clause 3(b), the part where it says "any third party". The key word is "any".
No, the key words are "third party". A person who does not posses the binary is *not* a party to this contract. "Any" is merely a modifier on "third party" indicating that it does not matter how the third party received the binary, i.e. receiving from someone other than the developer.
What about the rights of the people who worked on it in the past, with the understanding that their "compensation" would be in the form of others donating their time to continue improvements?
What "rights" or "understandings" of the past developers? Their right, their understandings, are defined by the GPL that they chose to license their software under. The GPL allows for payment. The only thing the GPL disallows is denying access to the source code.
...if only there was a similar situation we could use to predict how it might go.
... It depends on who has enough money to spare. Apple certainly has ...
And Microsoft did not when Wine got started?
Is there anything worth running?
Well the Xcode development environment is Mac OS X specific and unlikely to be ported to any other platform.
no? damn
Well, neither does the current Mac OS X. So it is fully compatible in that regard.
They should invest in 500 Mio in Wine. So you could ditch windows.
Apple already offers a solution for ditching Windows. :-)
the only way google did it in kansas city was by paying off local officials to allow them to put their fiber on the poles at lower rates than everyone else's lines. not going to work everywhere. ISP's and others are already suing kansas city for allowing this
Giving local officials some sort of deal seems to be a well established practice that has worked in many locales. See the cell phone industry. A cellular tower is opposed until the provider offers to put equipment to support local police and fire communications up there. I suspect that the initial opposition is just a gambit to get such freebies in some locales. I'd be surprised if such practices have not already been ruled on by the courts.
Well, according to a new report from Goldman Sachs, the price tag would be over $140 billion. Not even Google has that kind of cash laying around.
Apple could, they have $120 billion in cash. Maybe that is what they are saving up for ... becoming a national ISP. :-)