Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
Note: You can take 10% off all Slashdot Deals with coupon code "slashdot10off." ×

Comment Re:The problem with neural networks (Score 1) 44

No.
That's what the GP proposed.

For a human, a skill test is OK, because we already know he's a human, and cities are built around humans. We can expect him to behave in a certain way, and we kind of know his possible range of abilities and limitations, even if not in a formal way.

There's a reason why we require other things, like a minimum age, because being a responsible adult is precondition to the test.

What they are doing right now is different. Still a black box test, but much more comprehensive. They are going to gather so much data that statistics are going to start kicking in, bc they should be able to somewhat "prove" that these cars are safer than regular cars.

Comment Re:The problem with neural networks (Score 1) 44

That's just not true.
Humans, specially urban dwellers, are known to have a certain set of capabilities, in general.
Also, they are known to behave in a certain fashion, and to abide by certain rules.
For example, a human with tendency to kill everyone in his path, would just not be able to apply for a drivers license, he would be in jail, dead, or something similar.
That black box testing is only verifying very specific knowledge and ability. It doesn't do a great job at that, but its task is a lot easier than testing an AI from scratch. You could do that, if there was some "human like" validation test, you could take prior to getting a license.

Comment Re:The problem with neural networks (Score 2) 44

You make a very interesting point.
With automation, it's a lot easier for us to accept a given amount of understandable failure, than a much smaller amount of inexplicable failure. That might be a roadblock against some forms of automation.

In any case, there's also economics, which do like statistics, and will make you choose the strategy that fails less, overall. For example, insurance companies might favour driving algorithms that crash less often vs ones that crash a bit more often, but for better known causes.

Government

More Cities Use DNA To Catch Dog Owners Who Don't Pick Up Waste 171

dkatana writes: For many cities one of the biggest cleaning expenses is dealing with dog poop. While it is impossible to ask the birds to refrain from splattering the city, dogs have owners and those owners are responsible for disposing of their companion's waste. The few who shirk their duty create serious problems for the rest. Poop is not just a smelly inconvenience. It's unsanitary, extra work for cleaning crews, and in the words of one Spanish mayor, on a par with vandalism. Cities have tried everything from awareness campaigns with motorized poo videos, to publishing offenders names to mailing the waste back to the dog owner. In one case, after a 147 deliveries, dog waste incidents in the town dropped 70 percent. Those campaigns have had limited effect and after an initial decline in incidents, people go back to their old ways. Which has left many cities resorting to science and DNA identification of waste. Several European cities, including Naples and one borough in London, are building DNA registries of pets. Offending waste will then be tested and the cost of the analysis charged to the dog owner, along with a fine.

Comment Re:There needs to be a standard device (Score 1) 235

Nonsense.
Right now policemen are able to stop cars, no device needed.
Autonomous cars should be able to match current driver behaviour. There's room for improvement, but they don't need a better solution in all regards, to replace drivers, only to be as good as them.

Comment Re:Here's the thing though... (Score 1) 235

It's not really that hard for a bad guy to buy a cop costume. Humans can't tell them difference between the police and some random jackass. Also, if a guy is standing in the middle of the road signaling you to stop, you're gonna stop just to not run him over.

I think self-driving cars should be treated as taxis. Just like you can't expect your taxi driver to disobey a cop, nor can you expect your SDC to.

Good point, but that's a driverless taxi, not an autonomous car. An autonomous car would be a car you own and you command, that does what you ask it to do. Like a car, but instead of driving it, you tell it to go places, and override command _whenever_ you want it, not when it's lawful to do so.

I think that autonomous driving will probably be best applied to public transportation, because an automated taxi is a better taxi, while an autonomous car is a lesser car, after you take this kind of things into account.

Comment Re:Compatibility is not an unrealistic expectation (Score 1) 314

You say:

Sometimes code's behavior is the documentation.

Yes, that is the definition of undocumented software, when the only documentation is its observed behaviour.

But then you say...

Successfully reading and interpreting a file is one thing. Rendering its contents is another. The Office Open XML format, .docx, is open: ISO/IEC 29500.

No, it's not open. First, OOXML is a sanctioned standard, but it's not open. For example, there is no open reference implementation, only proprietary binaries.
Also, and most importantly, msoffice does not implement OOXML completely. Its small differences are what make compatibility a moving target.

About this...

So, the second to best solution is to just acknowledge msoffice is not compatible with other software, so either ditch it completely, or keep it completely. Half assed efforts are doomed to fail from the start.

That is a fine strategy if you are not trying to get people to convert to your app. Blowing off a valid user expectation hurts adoption.

But these guys were not trying to get people to convert to anything. Their job was to provide office productivity software for the city personnel. The hybrid solution was a bad idea, they failed because they tried something that is known not to work.

OpenOffice/LibreOffice guys can work on interoperability as much as they want. That doesn't automatically make it viable as a solution for a big organization.

Comment Re:Compatibility is not an unrealistic expectation (Score 1) 314

That, or you can RTFA.

The problem was unrealistic expectations. They went from an all msoffice operation, to a hybrid one.
msoffice is not compatible with anything else. You can migrate away from their formats, but you can't really interoperate with them without a lot of fiddling around.
That's costly, and wasn't accounted for in the original planning. Shockingly, it costed time and money.

I did read the article and you are completely mistaken. The problem is OpenOffice's failure at being compatible. If it paginates wrong an OpenOffice developer should fix that. If macros are missing an OpenOffice developer should add those.

OK, but a developer can't "fix" what is not specified. msoffice formats are not specified so that they can be implemented. So that's not something a oo developer would be able to fix by himself.

So, the second to best solution is to just acknowledge msoffice is not compatible with other software, so either ditch it completely, or keep it completely. Half assed efforts are doomed to fail from the start.

Being a government, they can ditch it completely if it makes sense for them. Where I live, there is a law that requires all gov data to be available in open formats, so it's even easier to comply that way.

Comment Re: Start open from the beginning (Score 1) 314

Governments can just do what's best for their citizens, and tell others to piss off.
It's a nice place to start.
Remember that governments deal with public data, so they should hold formats to higher standards than private companies. OOXML is just not open enough for government, partly due to the problems exposed in this article.
Also, they don't sell stuff, so they don't need their .pptx or .docx to be easy to open for their prospective customers. They can just offer docs and data in open formats.

Comment Re:Liars will tell the truth ... (Score 2) 314

That, or you can RTFA.

The problem was unrealistic expectations. They went from an all msoffice operation, to a hybrid one.
msoffice is not compatible with anything else. You can migrate away from their formats, but you can't really interoperate with them without a lot of fiddling around.
That's costly, and wasn't accounted for in the original planning. Shockingly, it costed time and money.

Comment Re:Shocking: a hybrid solution was expensive (Score 2) 314

Ok, I'll bite.
This whole story is about msoffice being _incompatible_ with anything but itself, and that issue costing the city a lot of money when trying to have a mixed environment.

This is just another example why their proprietary file formats should never have been approved as a standard, because in practice, they are not interoperable. Also, shows one of the the risks organizations face when using proprietary formats regarding access of information. Once you bite office, you are stuck with them for life, or for the life of your documents, whichever is longest.

Comment Shocking: a hybrid solution was expensive (Score 5, Insightful) 314

From TFA:

"We encountered several hurdles and dysfunctions around the use of specific features," Bruscoli says in the report. "What's more, due to the impossibility of replacing Access and partly Excel (various macros used on tens of files), we decided we had to keep a hybrid solution, using the two systems at the same time. This mix has been devastating," he adds.

They didn't replace MSOffice in the first place, they had a hybrid solution, which was costly, due to compatibility issues. They should have been able to know that beforehand. msoffice doesn't play well with others, it doesn't even implement any standard format. If you absolutely need to use msoffice in some spots, you should forget about interoperating, and just use msoffice everywhere.

Comment Re:Hero worship comes in all sizes (Score 1) 273

Sorry for taking so long to see your comment.

I don't embrace it, I just understand it.

I don't like the fact that assholes lead us, but they do. So if you really want to make a splash, you need to either become one, or just learn to work effectively with them.

There are several reasons why most members of the elite come out as assholes, one of them is that it's a trait that just works.

I think it's related to the prisoners dilemma. Society as a whole needs to be cooperative, but assholes thrive in a cooperative environment. The problem for us is that they also rise to power. So, being a good member of society (a good person) collides with leadership, or at least with rising to the top.

Then, there's the "class" issue. Asshole traits are a good indicator of you being part of the elite, so it's easy to cooperate with the elite, if you show you are part of it. That means that you can't just behave like a good person, if you want to make an impact, because the other guys at the top will think you are not one of them, and resist you. So you need at least a fine coating of assholeness, if you want to make an impact.

I don't know the right mix of asshole and good person, but I'm working on it.

Comment Re:Hero worship comes in all sizes (Score 2) 273

Doesn't seem right to me.
I don't like Jobs or Apple, but...

There are lots of engineers. As an example, Woz is as good as they get. He was instrumental to Apple's early success.
Nonetheless, Woz without Jobs did lots of great stuff, but nothing close to what Jobs achieved when no longer working with him. I think Jobs credit is well deserved. He did make things happen, and his own contribution did at least jumpstart the consumer smartphone industry, among other achievements. It was not a technical feat, it was all him.

I mean, engineers are good, but even when we are not cogs, you might think of them as the engines of innovation and technology. You need engineers, you need great engineers, but no single engineer makes enough of a difference to matter by his own. Maybe Jon Carmack may count as a notable exception, but most of the time, the tech guys are not the ones that make or break a company.

At the end of the day, the guys who steer the companies, seduce the customers and investors are the ones that matter the most and make the most difference. If you are an engineer and you think your vision is somehow better than the guy who runs your products, then you should become a product guy and leave engineering behind.

Comment Re:i haven't bought a car in a while... (Score 1) 252

I think the point is price.

Cabs are expensive, but most of the expense is paying the driver. Once you get rid of the worker, it gets a lot cheaper.
Also, with centralized control, routes can be optimized so the taxis are always driving and carying passengers.
It's not slightly cheaper than driving, it can potentially cost an order of magnitude less, and be faster.
Where I live, public transport is easily 5-6 times less expensive than driving, combining bus and cab rides, including the labor cost of drivers.

About keeping it clean, and accountability, we are now used to be identified always. The cars can have cameras, and even require an id for you to ride them (they won't be taking cash, after all). There would be abuse, but it would be close to trivial to punish that kind of thing.

If there's vomit in the car, the car should be able to detect that, and go to a cleaning station. In the event that you do get an unsuitable car, you can just reject it. You could even look at a stream of the inside of the car before it gets to your home.

Also, think about the carpooling possibilities. While people don't like sharing space with strangers, price can change some minds.

Top Ten Things Overheard At The ANSI C Draft Committee Meetings: (4) How many times do we have to tell you, "No prior art!"

Working...