Become a fan of Slashdot on Facebook


Forgot your password?

Comment Arleady problematic now (Score 4, Informative) 588

...when they finally go big time, given that the white lines currently are used to guide them on multi lane roads.

No need to wait for autonomous vehicle.
Current safety devices use it already:

- Lane Departure Warning:
vehicle uses the contrast of white lines on dark asphalt to guess where the lane is, and can either alert the driver (e.g.: Volvo cars) or correct course (e.g.: BMW) to stay in the lane. The driver needs to explicitly switch on the turn signal to tell the car that he indeed intend to turn the car.
No lines, not easy for the car to tell what exactly the trajectory should be. Whereas humans can more or less guess based on the surrounding and know where the "virtual lane" should go (and TFA's idea is that this guess-work will force drivers to be more prudent and slow down. My own feeling is that the first 2 weeks, the drivers will be watchful, then they'll get used it, and then everything will be back to normal)

- Forward Collision Avoidance:
vehicle have a forward facing radar that can detect other vehicle in front. So the car can see if the other in front breaks (when they are both in the same lane, i.e.: a traffic jam) and automatically slow down the cruise control (and in some car, resume driving once the traffic jam clears and the car in front starts again).
Also, the cars can detect incoming vehicle or vehicle that are on a crash course and prevent by applying breaks.
For that to work, again the car's computer need to have some basic idea of where lanes are. Other wise, there's a risk that the car will hit the break, even if the stoped/slower vehicle was in another lane, or the incoming car is in the other half of the road (like in TFA's case).

It seems similar to what i believe they did in the netherlands where they removed any distinction between the road and the pedestrian areas which apparently slowed down traffic.

...well at least, pedestrian and cyclist collision avoidance (more usually called "City Safety" by constructor, and currently slowly becoming a strandard option on most vehicle in europe), is entirely Lidar-based or shape-recognition based.
(i.e.: the car doesn't stop on its own because you're dangerously close to a pedestrian area or a bicycle lane, but because it recognised the object in front of you).
So at least *that* idea isn't disrupting existing safety device. But still...

I'm more proponent of some European city which have buried some of their highway network underground.

I don't think that forcing people to think about the security themselves by removing safety marking will actually work on the long term.
I strongly suspect that people will slowly adapt and get used to the missing markings, and start driving as carelessly as before.

If you think about it, large swaths of road miss markings, specially in developing countries. And those countries aren't exactly known for lower incident rates (though other reason, like vehicles to broken to be road-safe, missing driving education, etc. are other factors in play).

Comment (TFA citation) (Score 2) 504

Now, why not have them roleplay this (with an AI, or with a partner that would accept "playing" the subserviant) ?

Which is also shown as an example in TFA itself:

Interestingly, some AI assistants out there do cater to this sort of thing. CEO of Robin Labs, Ilya Eckstein claims there is a high demand for AI assistants that are "more intimate-slash-submissive with sexual undertones".

To each his own liking. And better to molest a virtual entity in roleplay that is designed to respond this way, than molesting a real person.

Comment Or the other way around... (Score 1) 504

Or you could consider the opposite:

Some of the people that you qualify "idiots" might have weird urges. They might *want* to degrade women even if they know it's bad.
Now, why not have them roleplay this (with an AI, or with a partner that would accept "playing" the subserviant) ?

Fed up that girls that you know / you (depending on sex) get constantly cat called?
Hey, why not build a special cat-call bot that the cat-callers can cat-call, and leave uninterested human females alone ?

Comment Intel Cores Schmores (Score 1) 136

Perhaps cores-schmores is one way to approach this? Lots of small cores with relatively slow clocks, as higher clocks tend to worsen power efficiency.

Which is also the road that Intel themselves pursue with Xeon Phi (the currently used descendant of their failed GPU).

I'm not discounting Intel's success with single-core performance per se, but I sometimes feel it's aimed at speeding up legacy applications

Yup, the drawback is that not a lot of current application are able to run on tons of separate threads.
Not only "legacy" but even applications recently produced or currently being produced.
But the architecture can have some success on servers, and some scientific workloads.

Comment Different types of material (Score 1) 165

While you're in checking mode, perhaps you could research uncountable nouns.

1. You don't add any value to the current conversation

2. Also, you're wrong : it's correct to use the plural :

I suppose you could take the totally opposite route and choose Shapeways or iMaterialize's rubber/elastic type materials

It might surprise you, but there are clearly more than one type of material that is flexible (e.g.: flexible nylon and printable rubber, just to cite the first 2 of the top on my head).
As the poster is referring to different types of material, rather then a bigger quantity of material, the usage of plural is correct.

Comment Batteries (Score 1) 223

There are many times during the year when I may need to drive 300 miles round trip. If it won't make it then it's a non-starter.

Two things makes this realistic:

- vehicle size: In a bigger truck, there's more room to store aditionnal batteries.
Whereas extending a Model S would necessitate fill the front and back trunk with additional batteries (increasing weight and killing potential cargo), on a truck you could realistically use more space for additional batteries while still having plenty of room left for cargo.

- also, electric motor are less complex and cheaper than internal combustion engines. Model X and newer Model S have two motors. European high speed train don't even feature a locomotive because *the whole train* is motorized - every single coach.
This increase either efficiency or (in P85D Tesla S) the peak performance.
So, to further the above point: On a "Tesla T" you could have the whole trailer motorized and with its own battery, dramatically increasing the potential range.
Tesla would need "simply" to design a flat-bed with motors and battery similar to its current dual-motor Model T/X platform, with the necessary point to attach a standardized shipping container for cargo.

- unlike a gaz tank, battery are swappable and it's a rather fast procedure.
Musk has always wanted to supplement Tesla's network of free super-charger, with a network of (paying) battery swaps.

In other words, planty of opportunity for electric trucks in some point in the future.

So by applying the same approach and

Comment Technology vs. Implementation (Score 1) 44

>The main advantage of bitcoin and other crypto currency protocol, is that there isn't a single entity in charge of the transactions, there's not a single point that you can block/ban.

Wrong and wrong. has enough power to control the chain. Around 10 people basically control BTC

I was speaking about the general concept of the design.
Not the peculiarities of the implementation.

The bitcoin (and other cryptocoin protocol) are designed to eschew the need for a central entity (compared to other exchange protocols and platforms that need a central authority and couldn't work without one). By design, bitcoin doesn't need one, because by design it distributes the information across the whole network. And thus by design it CANNOT be anonymous. At best, it's pseudonymous (there are no Real Identities, so an user could seem anonymous at a quick glance). At worse one need to use complex coinmixing operation to manage to maintain anonymity.

The fact that a huge a part of the network is at the hand of few key player and in partice there's an oligarchy controlling it, isn't a result of the design (the protocol is designed without the need of an authority) but of how things have evolved practically for the current implementation:
BTC overly rely on Proof-of-Work (which at some point of time was important to attract new players to grow the network), and beyond that has focused on a specific PoW - a variant of Hashcash - that is computationnally simple, scale dramatically fast (each new generation of hardware completely leaves the older one in the dust) and thus in the long term works best for those with plenty of cheap electricity and cheap access to electronics.
Of which China has plenty (they have Three Rivers Dam, and they are the one who make the electronics for everyone else). So of course they'll end up being strongly advantaged, and few key chinese will hold much of the network's hashing power, and therefor would work as the de facto leader.

That doesn't change the design was to be without the need of a central authority, but be distributed accross a whole network (of which they simply managed to hold the most).

Or in short: The general idea behind the technology is interesting. The current bitcoin network is slowly turning into shit. But one doesn't change the other.

Comment Re:Already here (Score 1) 412

You can still better yourself and get a better paying job, just no free cable TV, smartphone, etc.

I think "smartphone" is a poor example on your list, as it is slowly becoming a critical piece of technology to be able to do anything.
Just like "a computer and a working internet conection" has been in recent times.
It's not just a piece of entertainment (like a TV), but a critical piece of technology to get access to maps, tools of communication (both voice and text messages), reading mail while on the go, getting information, etc.
There are lots of jobs where you basically need a smartphone to be able to work (random example: Uber driver*).

*: Though it's a bad exemple for this discussion, as we're currently speaking of Europe and to work as a driver there, you need a professional driver license, special profressional insurance, etc. these aren't cheap and thus working as an Uber driver isn't an entry job that cou can do when a smartphone and a car are your only pessessions in this world.

You buy what you can afford on your income rather than living above you means.
You want a better lifestyle? Do what the rest of do and EARN it.

The problem is : what happens with people who have always worked to be able to earn the lifestyle. But suddenly aren't able to work for reasons external to their will (e.g.: sickness/accident)
They are willing to work. The have worked up until now. They just suddenly can't anymore.

Comment Salary (Score 2) 412

So once you're in that "basic income" system of yours, I guess you're stuck living in some ghetto and would have no way of getting out of it.

1. It's European countries you're speaking of. Here around, what you call "some ghetto" are way nicer place than any of you ghettos on your side of the altrantic pond.

2. They idea is: "this buys you minimal living accomodation in the more modern parts of a big cite / or in a really small village lost in the back country, now it's up to you to earn anything more you would need to be able to access anything more that you would want"
Deciding to get a paying job is basically *THE* way of getting out of it.

Comment FFMPEG (Score 2) 133

After the FFMPEG fork is there a Linux distro that still uses FFMPEG & Mplayer?

Firstly: Not all distro switched to avlib.
Some simply decided to stay with ffmpeg (e.g.: opensuse never switched at all)
Some changed their opinion back (e.g.: Debian went back to ffmpeg after a while)
This is mostly to avlib never really being a good an active fork, and didn't manage to attract most developper to it.
(Unlike to which most developer migrated after the fork from

Since then the problematic leader of FFMPEG has decided to step down,
avlib has merged back to ffmpeg
and distro are back to- / or are still using- ffmpeg again.
And the guy is now a contributor. He still writes code for ffmpeg, but he's not having a final say on everything and thus fighting with everyone to have his "one true vision(tm)" imposed.
He has fully realized that his character clashes with some in the community, has seen the disastrous result on having avlib forked (the linux ecosystem split across two different forks, none of which becomes a clear leader and each laging behind the other on some important features), and decide therefor to step down for the greater good of the community.

source: Phoronix

Comment Volvo AREN'T autonomous (Score 3, Interesting) 229

Please don't put people in it if autonomous.

The Volvo aren't autonomous in the sense that they don't handle the actual route.

Volvo mainly use their sensors (though it's camera + lidar + radar, just like on autonomous cars)
to detect possible objects that could collide with the car and break and/or sound an alarm.

Note that the driver can still override by slamming the gas pedal. (People want to be able to have the last say).
But if the driver doesn't do anything, the car will automatically slow down and stop before hitting the car/pedestrian/whatever in front.
(And also, resume driving if the car in front starts moving again. That's a very useful feature in a traffic jam. Though if the Volvo has stopped for a longer period of time, it asks a confirmation from the driver (button or gas pedal) just to be sure to have the drivers' attention.
After all, its NOT an autonomous car, and the driver is still responsible, so it would be better if the driver hasn't dozed off during the stop).

Ok Volvo, see if your car can drive this

Some of the feature of Volvo car are already useful in these situations.

Again, Volvos aren't autonomous, it's NOT their job to actually drive though this kind of hell.


The lidar and radar will correctly whatch for anything the car might crash into.
The volvo will correctly stop before crashing into incomin vehicles or against the mountain (due to too narrow space for crossing).

The camera tracks the road and can sound an alarm if the driver risks quitting the path.
(Though unlike other brands like BMW, the Volvo won't correct the course by itself. It just sounds an alarm when detecting that the driver was swerving away of the path and either:
- hopes that the driver will wake-up correct and course
- of the driver will turn on the turn signal, because the driver was actually swerving away from the current lane on purpose - he/she wanted to change lane, but without the turn signal, the car couldn't know it and sounded an alarm anyway. Of course that last one applies to changing lane on a multi-lane highway.
Not trying to stay in path in the kind of hell like this mountain "goat-path-except-there-are-truck-on-it" from your terrifying example).

Comment Alpha car software analogy (Score 1) 111

so the cars are in an early alpha test, to use a software analogy?

Yes, and this alpha-car softrware analofy works MARVELOUSLY well, FULLY AUTOMATICALLY, requiring absolutely ZERO USER INTERVENTION / OPERATOR SUPERVISION...
...on the small test dataset.
Which is exactly a 1mb sized file.

But we're sure that one day we'll be able to plug this software analogy into the data center to service gigabytes of data per second.

Comment Not WRONG but POORLY UNDERSTOOD (Score 1) 251

The poster isn't exactly *wrong*.

He just have poorly understood and reinterpreted in own words, an actual problem that does exists for real:
Some qualcom chipset have the modem inside the main SoC and that modem works as a "sort of northbridge" for the SoC.
The modem is in charge of handling RAM, audio hardware, GPS, etc.
That modem, for legal reason - runs a 3rd party firmware that is provided by the phone service provider.
Android runs on a CPU core that is client to this modem to access the phone resources.

See my other answer in this thread.
And see the Replicant wiki.

Comment Modem firmware, instead of SIM itself. (Score 1) 251

It's not the CPU core and memory that is inside the physical SIM.
As I have explained in my other answer in this thread, it's the modem part.
The modem - which for legal reasons runs a 3rd party closed source firmware provided by your service provider - of several Qualcomm chipset works as "sort of northbridge" to the chipset.
The modem (and its 3rd party firmware) is in charge of several critical parts of the phone, which may include RAM, audio hardware, GPS, etc.
Android runs on a CPU core that function as a client to this modem an accesses everything thourgh it.

Replicant have complained about this in the past and documented in their wiki.

Comment Not the physical SIM card itself (Score 1) 251

It's not the physical SIM-card itself.

See my other answer in this thread.
And see the Replicant wiki.

On some chipsets by Qualcomm (which are extremely popular) the *modem part* serves as a northbridge to the chipset.
It handles some critical component like RAM, sound hardware, and OS is running on a CPU core that is a client to that.

And for legal reason, the entity responsible for the code running both on the physical SIM card it self and running in the modem firmware is the service provider.

Regard TFA, that means that even if Google decide to say "Screw you!" to NY and CA legislation, the phone service provider is just one governemnt letter/order away from getting all your data.
(because, remember: all your data is on a flash medium that is directly plugged into the modem running the service provider's firmware. Your Android is running on a CPU core that is a client to this modem).

Slashdot Top Deals

Computers are unreliable, but humans are even more unreliable. Any system which depends on human reliability is unreliable. -- Gilb