Comment Re:I'm so excited (Score 1) 116
I have not claimed it is difficult. You did not correct my reading.
I have not claimed it is difficult. You did not correct my reading.
Nothing limits you to one bit-per-second per baud. 9600 bps modems were, IIRC, 2400 baud with 4 bits per Hz. (Higher than that it got a bit shady because they started optimizing for being encoded in a digital phone line).
VDSL2 goes up to 32768-QAM, which is 15 bits per symbol. I do not know whether any actual phone lines exist with a sufficient signal-to-noise-ratio to make that coding useful.
So you know that Huawei devices are back-doored because they are knock-offs of back-doored devices from Western manufacturers?
I must have read that wrong somehow.
I would like to congratulate the ISEE-3 team on helping avoid a man-made impact with our moon.
I am glad that the satellite was saved. However, why is it good to avoid man-made impacts with our moon?
The Star Trek transporters are clearly murder factories. However, a proper quantum state teleporter would not allow duplication, and it is possible that the consciousness would travel too. Then again, it might not, and there is no way to tell.
What good does it do me that there is a copy of me with my memories on a different planet? If my consciousness does not go to the copy (and how would it?), it is rather pointless as well as creepy.
The point you miss is that humans also won't notice the loss of a 20 or 40 milliseconds of data as long as the receiving end does something reasonable (e.g. replicating last received packet).
This is not true. Packet loss of less than 1% is detectable by actual customers with typical VoIP phones and the Alaw (or ulaw) codecs. It would be very handy to handle links with 0.3% packet loss, but that requires switching to codecs with built-in PLC, and transcoding is expensive.
In several markets you can buy petrol which is supposed to reduce losses in the engine. It has the same octane rating as regular petrol.
Whether it actually works I do not know. I like that fuel is a commodity, I resent the recent attempts at changing that. I wonder if I will be able to go to charging stations with special long-running electrons in ten years...
It really isn't that hard to get very close to the ideal fuel consumption figures, you just have to relearn how to drive instead of going full-throttle/full-brake all the time.
You are driving a car with non-faked figures. Of course you can get close to the official figures.
Relearning does not help you with most modern cars. They cannot physically achieve the numbers listed.
I don't want your car ruining my air.
4500 minutes at 3G quality is on the order of 500MB. Surely upping your bandwidth cap by 500MB is insignificant when you pay for 4500 minutes.
Error correction does not introduce anything like 400ms latency. 20ms lets you conceal the loss of one packet perfectly. 40ms allows you to lose two in a row. No humans will be able to detect 40ms latency.
Modern wide-band codecs use on the order of 50kbps for really high quality voice. If you use 50kbps continuously, you have added 16GB to your data cap each month (or 32GB is you transfer full duplex). Useful perhaps, but your phone would be constantly busy and battery life would suck. What are you going to transfer at 6kBps per second anyway?
100 ms latency is perfectly fine for voice. VoIP from Europe to China with 400ms is surprisingly usable, most users do not notice the walkie-talkie effect. Anyway, it is trivial to do error correction for VoIP with far less than 100ms latency. The easiest solution is to include a digest of the previous packet in the following packet -- if the previous packet arrives later than the following packet or not at all, use the lower-quality data in the following packet. Spread the information across multiple packets and you will be able to handle multiple packet drops of course.
I wish Asterisk would come with a simple-stupid PLC codec mode where it simply appended the previous packet contents in their entirety to the following packet. It would double bandwidth, but packets-per-second would be unchanged, and 128kbps is nothing on most modern lines. The added 20ms latency would be unnoticeable.
Voice traffic is not worth measuring. It just does not take up bandwidth at all.
Today is a good day for information-gathering. Read someone else's mail file.