I'm sure you're not the first to think of that. IRC Bosch is currently the market-leader for car-radar solutions, they
do have a fairly good track record of making reliable car-parts.
Digital Signal Processing goes a long way. With multiple antennas beam-forming is possible. Current systems send out chirp signals.
Different cars will not be locked in phase, so that will automatically reduce interference also
They started with getting it to work on one distribution (on of the more popular ones), they will get it to work on others.
The articles describing how the worked with graphics card manufacturers to improve performance on linux suggests
that their engineers are quite skilled, but only human, so they cannot do everything at once.
Not really, since the government can (and does) order wiretaps, the google glass now allows the wiretapper
to see what the target sees.
Wiretapping is strictly for governments (well, and criminals).
that they never get close to the bang-for-buck that the US gets.
Quite true, but what happens when China spends, say, twice as much on R&D ?
They will overtake US at some point, if the current trends in both US and China continue.
The one thing that trips me up with python is the link to C++. Cython seems a bit hakish, ctypes
requires plain C.
Plus, I'd like to be able to move algortihms from high level to low-level once profiling indicates that
it's a performance critical part. That's easier if both high- and low-level are the same language.
Keeping everything in one language is simple. Once a build-system is set-up, re-compiling+running a C++
program takes about as much time as starting a python script.
Both C++ and python applications can be fast or slow to startup, determined by how many dll's/dependencies are loaded.
So for a quick test, I'd prefer python, for anything serious I use C++.
I've always wondered about the dynamic typing:
How does that hold up in large projects, with dozens of programmers ?
For me, the static typing of c/c++ catches a lot of errors/typos in function call arguments.
On the other hand, it's the only mainstream language I know that supports both very low-level and very high-level programming style.
This can be a real plus for compute intense signal processing, were a small minority of the code really requires low-level implementations. Being able to mix that with high-level abstractions (e.g. linear algebra factorizations) can give both efficient and maintanable code.
It used to be that stress-testing HD's with random disk access for one day could flush out a lot of bad ones.
The ones that did survive tended to last many years.
It's a tricky thing with RAID drives. If you happen to have bought a 'bad' batch, chances that more than
one will fail before you replace one are pretty high. So testing makes sense sometime.
A while ago, google published some research to show that drives do not fail randomly, but in clusters.
Making RAID a bit more susceptible to data loss than one might expect.
The square waves part is really wrong.
That's how the data is often displayed on a screen, but really not how it is played back. There is a low-pass filter in a D/A converter which guarantees that the output is smooth (and band-limited)
Yes it does make sense.
He's talking about dynamic range compression though, not digital compression as in zip-files. The same word for 2 totally different concepts can easily cause confusion
Well, you could test.
wipe the disk, write a known pattern, pull the plug,
dump to screen the last byte that is written according to NCQ,
re-plug, read.
It may be surprising, but the icy/rainy slippery roads are a walk in the park for a computer controlled vehicle.
Current anti-lock/stability controls do show excellent reliability already.
It's reacting to unknown/unexpected conditions which can be tricky:
road constructions with messed up road markings, cities with unpredictable pets/small children.
Those I'm far more worried about
MMm, both the German ICE and the French TGV work just fine, allowing one to quickly travel
large distances in comfort.
Comparing that to Great-Brittain, it seems that railway systems fare better in 'socialist countries
with strong government.
Slashdot Mirror
I'm sure you're not the first to think of that. IRC Bosch is currently the market-leader for car-radar solutions, they do have a fairly good track record of making reliable car-parts.
Digital Signal Processing goes a long way. With multiple antennas beam-forming is possible. Current systems send out chirp signals. Different cars will not be locked in phase, so that will automatically reduce interference also
Radar (around 60GHz) is quickly becoming more cheap. It's antennas can be much smaller (and cheaper) than a LIDAR setup.
A real man drives a stick shift, self-driving cars are for whimps! (Yes, I'm european)
Perhaps their engineers are not that skilled?
They started with getting it to work on one distribution (on of the more popular ones), they will get it to work on others.
The articles describing how the worked with graphics card manufacturers to improve performance on linux suggests that their engineers are quite skilled, but only human, so they cannot do everything at once.
Not sure if you're serious, but it's at bit hard to google for that, even with the chocolate factory's save-search enabled.
Well, ethiopians for some reason perform remarkbly well at marathons. Is that rasict ?
Not really, since the government can (and does) order wiretaps, the google glass now allows the wiretapper to see what the target sees. Wiretapping is strictly for governments (well, and criminals).
So if you pay with cash, you pay the fee twice: Once at the ATM, once in the store. When paying with a debit card, you only pay it at the store.
He's talking about PS2, you're talking about the PS3
Don't forget to wipe out the BIOS, or disable netbooting... Security ain't easy
But the boat still sunk, right ?
that they never get close to the bang-for-buck that the US gets.
Quite true, but what happens when China spends, say, twice as much on R&D ? They will overtake US at some point, if the current trends in both US and China continue.
The one thing that trips me up with python is the link to C++. Cython seems a bit hakish, ctypes requires plain C. Plus, I'd like to be able to move algortihms from high level to low-level once profiling indicates that it's a performance critical part. That's easier if both high- and low-level are the same language. Keeping everything in one language is simple. Once a build-system is set-up, re-compiling+running a C++ program takes about as much time as starting a python script. Both C++ and python applications can be fast or slow to startup, determined by how many dll's/dependencies are loaded. So for a quick test, I'd prefer python, for anything serious I use C++.
Objective-C.
I've always wondered about the dynamic typing: How does that hold up in large projects, with dozens of programmers ? For me, the static typing of c/c++ catches a lot of errors/typos in function call arguments.
On the other hand, it's the only mainstream language I know that supports both very low-level and very high-level programming style. This can be a real plus for compute intense signal processing, were a small minority of the code really requires low-level implementations. Being able to mix that with high-level abstractions (e.g. linear algebra factorizations) can give both efficient and maintanable code.
It used to be that stress-testing HD's with random disk access for one day could flush out a lot of bad ones. The ones that did survive tended to last many years. It's a tricky thing with RAID drives. If you happen to have bought a 'bad' batch, chances that more than one will fail before you replace one are pretty high. So testing makes sense sometime. A while ago, google published some research to show that drives do not fail randomly, but in clusters. Making RAID a bit more susceptible to data loss than one might expect.
The square waves part is really wrong. That's how the data is often displayed on a screen, but really not how it is played back. There is a low-pass filter in a D/A converter which guarantees that the output is smooth (and band-limited)
Yes it does make sense. He's talking about dynamic range compression though, not digital compression as in zip-files. The same word for 2 totally different concepts can easily cause confusion
The same holds for these apps. Same difference.
MMm google gives 112km/h, which is a zippy commute... try googling for: 0.000000032 parsecs per leap year in km/h
Well, you could test. wipe the disk, write a known pattern, pull the plug, dump to screen the last byte that is written according to NCQ, re-plug, read.
It may be surprising, but the icy/rainy slippery roads are a walk in the park for a computer controlled vehicle. Current anti-lock/stability controls do show excellent reliability already. It's reacting to unknown/unexpected conditions which can be tricky: road constructions with messed up road markings, cities with unpredictable pets/small children. Those I'm far more worried about
MMm, both the German ICE and the French TGV work just fine, allowing one to quickly travel large distances in comfort. Comparing that to Great-Brittain, it seems that railway systems fare better in 'socialist countries with strong government.
And if I read the pages correctly, NASA's probe still works, while India's stopped working after a year. thank you, come again...
Transmission is nice for small servers (it has a web-interface) qBittorrent is good for the laptop/desktops