Forgot your password?
typodupeerror

Comment: Re:This seems like a good time to meniton these (Score 1) 168

by udippel (#47690165) Attached to: Processors and the Limits of Physics

Yep, 'they' do. Where 'they' includes me. I was afraid that's what you'd thought; and that's fundamentally wrong. Propagation delay and processing time doesn't render a circuit asynchronous.
Following your logic, any circuit would be asynchronous, if only on the propagation delay induced by any non-infinitely short wire. Then we'd say that the whole world is asynchronous; and that'd be it. When we use these terms, however, it is common understanding that a synchronous circuit is 'guided', clocked, by a central clocking device. And all sub-circuits are controlled by that same device. Likewise, an asynchronous circuit is controlled by the actual occurrence of the signal, without a predefined time frame. Respectively by a pre-cursor (header) in the signal path.
One could as well focus on the number of signals: a synchronous circuit has a clock signal and payload signal(s), while the asynchronous circuit has only one signal: the payload. Eventually with a header.
 

Comment: Re:can't cross chip in one clock. big deal. (Score 1) 168

by udippel (#47688223) Attached to: Processors and the Limits of Physics

Sure, throughput is what matters most for operations you can parallelize. However, as Amdahl's Law cruelly reminds us, there's always parts of the problem that remain serial, and they'll put an upper bound on performance. You can't parallelize the traversal of a linked list, no matter how hard you try. You have to invent new algorithms and programming techniques. (In the specific case of linked lists, there are other options that trade space for efficiency, such as skiplists.).

Not for forget, Amdahl's law is purely calculation; not speculation; like what Moore's law is. And you are right, certain items simply remain serial.

Comment: Re:Piheads are like the guy with a Hammer... (Score 1) 427

by udippel (#47636657) Attached to: Ask Slashdot: Life Beyond the WRT54G Series?

Seconded!

I've been running a Soekris Net48-something for the last 10 years. Porfect, with m0n0wall. In 2014, however, 100 Mbps and a Geode 486 *are* getting long in the tooth.

All those cheapo single-NIC versions cannot convince me; like the original WRT54.

Also, I can't make out why a dual- oder quad-ARM should cost a couple of hundred bucks? They come with design, casing, 3G, touchscreen, batteries, and stuff at below 100 as mobile phones.

Comment: Actually, it does not 'turn out'! (Score 1) 275

by udippel (#47636635) Attached to: Long-Wave Radar Can Take the Stealth From Stealth Technology

Hey, this is not news! Of course, if you want to absorb waves with dampening material, you'd have to have a depth of the material of around a quarter of the wave length. This is know in acoustics as well as RF engineering.
So the news would be that some parties are developing radar systems with combined low- and high frequency excitation.

Comment: Re:What?!? (Score 1) 928

No rights were violated unless Southwest Airlines recently became government owned.

Also observe that nobody stopped the man in question from tweeting anything, it's just that the airline after reading the tweet decided it didn't want to transport this person. And that is fully legal.

4?? Insightful?? Who's the mods who are doing stuff like that; stuff that irritates?
Megol could be an employee (of SWA); a lawyer he probably isn't. Don't know about US law, but where I reside, this is not the case. Once the airline takes the booking, it also takes an obligation to transport you. (I really can't believe that the law in the land of the free would be that free!)
Megol has an ID that could be set up for astroturfing. But the 3+ mods who mod this 'Insightful'? Which allowances helped them to cast common sense aside?

Comment: Re:You read it here ... (Score 1) 435

by udippel (#47470009) Attached to: FBI Concerned About Criminals Using Driverless Cars

I *think* you got the message: I am worried about the future bug-free, perfect, software.
The one that gets via Internet -of-things the number, age, gender, medical record of the passengers in another car/bus, to minimize casualties. What a brave new world, when we start to steer vehicles in problematic situations into a state of minimal loss to society.

Comment: Re:You read it here ... (Score 1) 435

by udippel (#47469957) Attached to: FBI Concerned About Criminals Using Driverless Cars

I imagine it'd be program to slam on the breaks and stop, minimizing the damage. Don't give me that bullshit about swerving to the side. In that case it's program to kill whoever is in violation of it's right-of-way.

Did you get my argument? Minimizing whose damage??

Really, you think that the car is going to have some sort of morality judgement function. No.

Yes, it will. Not that I would like that, but it will have to be mandated. Imagine a car manufacturer who does not build in such a device! The hell will be sued out of him, for not programming a function minimizing human casualties. Because it becomes possible with this type of technology.

Comment: Re:You read it here ... (Score 1) 435

by udippel (#47469895) Attached to: FBI Concerned About Criminals Using Driverless Cars

It's a numbers game, and the very rare cases you are excessively concerned about do not even come into consideration.

It's a numbers game, you write, and further up, to consider own safety first. I am not clear which you actually want to reason about. If it is just 'my survival first', a single passenger would be saved to the detriment of a group?
How does this come across? The car you are sitting in deliberately runs down 5 people to save you? So, in that logic, a pedestrian, e.g., is less valued than someone in a car?

Comment: Re:You read it here ... (Score 1) 435

by udippel (#47469859) Attached to: FBI Concerned About Criminals Using Driverless Cars

Humans make worse decisions in those situations. Hell, many of those situation wont occur in the first place.

The issue is you're afraid of not having the illusion of control you now have.
In the situation you describe, humans won't have time to make a decision. If there would be enough time for humans to make a conscious decision, then that's enough time for a computer to have negated the issue.

I do agree that in average, the mortality on the roads will go down. Humans make worse decisions. Though, in most such cases they don't make any decision at all. In the case of the driverless car, the decision is, as I wrote, premeditated. And when it is the choice to either run the car into a group of people or off a cliff? I wouldn't want to kill the people, though I also wouldn't want the breaks of the car I am sitting in, to not be activated based on a quick decision by the algorithm that controls activity in cases of lost causes. And deciding to let us off the cliff instead of endangering a crowd of 15.

Comment: Re:You read it here ... (Score 1) 435

by udippel (#47469807) Attached to: FBI Concerned About Criminals Using Driverless Cars

Humans are sometimes faced with those decisions, too. The reality is that there are some solutions that don't have any good answers -- just a bad answer and a worse one, and we can argue about which is which.

For every one of those kids/octogenarians the computer decides to execute, it could be that there will be 100's of people who get to live because computers took over the driving. So your "look at all the blood" argument is a little misleading.

Exactly. It will be a deliberate, or even conscious, decision to - as you correctly wrote - execute either. Humans make more mistakes than a driverless car; no doubt. Though, the argument who to kill is not left to some panicky reaction of a rudderless driver with not much of cognitive capability at that moment. Instead, it must be programmed ante mortem.

Nature, to be commanded, must be obeyed. -- Francis Bacon

Working...