If you consider SVN, GIT, or any code optimized source control, don't forget that many business users rely on the last date modified attribute of the file to determine versioning. Many source control systems don't care and don't store file modification dates (they store check in dates instead). This can be a show stopper. While as technical people we want to force the concept of version numbers on folks, it is just not in the culture of many business types.

Sweet. Please consider taking this concept back to the team. On the unlock screen have three unlock codes: 1) The normal unlock code, 2) A "limited" unlock code that would allow access to a limited set of applications (and perhaps make the device look like it was mostly empty), 3) a "wipe" unlock code that wipes the device (or nukes the encryption key.) An additional "distress mode" unlock code could be useful as well - this mode would start audio and video streaming to some off-phone storage. The key feature is that these special modes are activated by entering an alternate unlock code (or pattern). That way if the officer (or adversary) is asking you for the code and entering it on your behalf, you would still be able to activate these features.

In that case the call center would have to have a way to route any given audio clip to the overhead public address system, because you know that certain audio clips just have to be heard by everyone in the room.

More likely the recognition software assigns a confidence score to its recognition. High levels of confidence for a command are just executed. Low levels of confidence are probably filed for human review - probably not in real time, but for later analysis to tune the software.

Also a Frontier customer in Oregon (as a result of the sale from Verizon). I have had no issues with service since the sale. I have both the FIOS and hard wire line service and they both work great.

Fraud is already illegal. Perhaps we should enforce that.

Or a device that broadcast a very specific non-random set of keystrokes. For example you could send the keystrokes to open up a command window followed by the keystrokes to download and execute malware. You could even send the keystrokes to type in the source code and compile the malware or a malware bootstrap process.

I make only one resolution each year - That is to make no more resolutions until next year.

How would you prevent malicious use of the "do not send to the source/destination" packets?

While this is a nice list, are you prepared to to pay a lot more for every ticket in order to get the consistency that you want? Airline pricing is the way it is because it allows the airlines to make a mostly predictable profit under very thin margins. If the airlines are forced to adopt your list, they would still need to make a profit and would have to significantly increase ticket all prices to ensure their profit. This is because the current pricing model is vary dynamic (daily / hourly price changes) based on the expected number of passengers. You are suggesting the the pricing model no longer be dynamic. The airlines would not accept the risk of losing money and would have to increase the (proposed non-dynamic) prices.

There are plenty of credit cards out there that have no annual fees. If you pay your bill on time and in full each month, you don't have to pay any interest or other kinds of fees. If you can't manage to remember to pay your bill on time, you can pre-pay your credit card. You might have other reasons to avoid the general advice of using credit cards over debit cards, but "extortion fees" isn't really a valid reason.

No, you are just old.

I am glad someone got the joke. I considered talking about two modes of operation ("line of sight mode" and "make line of sight mode") instead of error correction algorithms, but I thought the error correction thing was more subtle and humorous. I guess it was too subtle. I don't plan on quitting my day job.

Not if you have the right error correction algorithm. If packet transmitted successfully, send next packet; otherwise increase power and try again. Repeat as necessary.

In concept making the .NET framework open source sounds cool. But, does making it open source mean that I can make a change to the framework, recompile it, distribute the binary framework along with my dependent application, and expect that someone else can just install my version of the framework and be good? Or does it mean that if I want to distribute a modified framework, I have to go through some sort of code signing process in order to allow it to run on someone else's computer? What would this signing process look like? In addition the assembly loader in .NET makes certain assumptions about version numbers of assemblies. It will be interesting to see how it works when due to multiple development paths, core assembly version numbers are not necessarily sequential / increasing over time.