Slashdot stories can be listened to in audio form via an RSS feed, as read by our own robotic overlord.

 



Forgot your password?
typodupeerror

Comment: Useful? Maybe if you could disable this 'feature' (Score 2) 191

by Behrooz (#48086823) Attached to: Belkin Router Owners Suffering Massive Outages

It might be useful to have a way to disable this 'feature' on the client side.

The bad? There isn't.

The good? This 'feature' already broke connections for anything going through the campus NAC even before their heartbeat server crapped out. SOP for any Belkin-involved problems became "Belkins break RFC2616, they are officially unsupported, go return it and get something that doesn't suck." ...so there aren't any of them still in use to be broken today. Yay!

Comment: Arbitrarily breaking HTTP is a bad idea. Who knew? (Score 4, Interesting) 191

by Behrooz (#48086755) Attached to: Belkin Router Owners Suffering Massive Outages

Entertainingly enough, I've run into this issue before. You will encounter the same issue when trying to connect the affected Belkin routers through the Cisco Clean Access NAC here (AKA Campus Housing), because devices are quarantined in the VLAN ghetto until successfully authenticated and associated.

So, these terrible, terrible Belkin routers try to phone home, and when they are unsuccessful they redirect all HTTP requests to the router's administration page. Since sessions are required to authenticate via HTTPS, there is no way to login. Extensive investigation revealed no way to disable this behavior on the client side, SOP for anyone calling with connection problems involving a Belkin router became "Officially unsupported. Return it and get something else that isn't a Belkin."

I am beyond pleased that this incredibly foolish decision on Belkin's part has come back to bite them in general, and hilariously entertained to see that Belkin's temporary workaround was effectively "spoof DNS traffic to heartbeat.belkin.com to a server on your local network that will pingback to fix your ISP's broken clients"

Comment: It's called 'bundling'. (Score 3, Informative) 341

by Behrooz (#47762959) Attached to: Comcast Tells Government That Its Data Caps Aren't Actually "Data Caps"

I get the "bought" part, that is after all how lobbying works (it's not a secret), but how does one "sell" a politician? Do you mean that political parties are pimping out their people?

It's called 'bundling'., where existing wealthy donors who have already contributed the legal maximum 'sell' the candidate to their friends and business associates, effectively leveraging their personal connections and access to shepherd more funds to the campaign.

What has more political clout than one maxed-out contributor when it comes time to make policy? A fucking cartel of maxed-out contributors.

Given that your average congresscritter spends ~20% of their working hours trolling for contributions just to have a decent shot at getting re-elected, you can imagine how influential successful bundlers are.

Makes you wonder just how much we'd save by spending a couple billion a year on public financing of elections.

Comment: Why worry about skyscrapers? (Score 4, Insightful) 191

by Behrooz (#47743703) Attached to: Slashdot Asks: How Prepared Are You For an Earthquake?

I'm not sure I'd want to be in a skyscraper in Memphis or St. Louis during a replay of the New Madrid quakes of 1811-1812,

Almost certainly safer than anywhere else. Skyscrapers are pretty much universally steel-framed structures which are relatively resistant to seismic loading, subject to stringent building codes, by definition need massive foundations driven to a solid base, and already need to resist dynamic wind-loading forces with resonance effects. Even mid-rise 6-10 story buildings are likely to be quite safe given the inherent seismic benefits of steel-frame construction and attention paid to building codes in the USA.

Has any modern skyscraper ever experienced significant structural failure resulting in loss of life as a result of an earthquake? Ever? Even in areas known for less-than-enthusiastic enforcement of building standards?

Comment: Re:How accurate is the Eurekalert article? (Score 1) 116

by Behrooz (#47687583) Attached to: Fukushima's Biological Legacy

The difficulty with any study of low-level radiation doses on exposed organisms is that ecosystems are messy and complicated, and the actual low-level biological mechanics for low-dose exposures are entirely conjectural, so observed effects have a very, very low signal to noise ratio.

Setting up controlled studies with a large enough scale to make statistically significant judgements greatly exceeds available resources for researchers in the field, leaving statistical analysis of effectively uncontrolled real-world populations as the only option.

So, the conclusions are only as reliable as the observations and the statistical analysis. This becomes educated guesswork, and for the most part educated guesswork based on the theoretical model that is being tested.

It's all very well and good to throw out statistics like:

Many other cell types and tissues have been shown to be affected by Chernobyl contaminants. Møller, Bonisoli-Alquati, et al. (2013) demonstrated that the frequency of visible tumors on birds was significantly higher in radioactive areas, presumably reflecting elevated mutation rates in somatic tissues. Visible tumor rates in birds from Chernobyl were in excess of 15/1000 birds while tumors have never been observed in Danish populations despite extensive surveys (0/35000 birds observed) (Møller, Bonisoli-Alquati, et al. 2013). ...unfortunately, the former USSR as a whole exhibits levels of persistent organic pollutants several times greater than observations in even industrialized areas of Western Europe. POPs are rather easier to study, and are definitively linked to tumor formation.

Attempting to control for various effects in real-world populations is a black art, often practiced and seldom practiced effectively. Often, before you can even start to evaluate the reliability of an article, you'll need to jump several citation links back just to see what assumptions a study is based upon. ...and that is why we still lack conclusive evidence about any long-term negative effects of low-dose radiation exposure.

Comment: You want references? LNT isn't a useful model. (Score 1) 230

by Behrooz (#47495751) Attached to: EPA Mulling Relaxed Radiation Protections For Nuclear Power

The difficulty being, your references are estimates based on what dose threshold?

Well, you have to go three citations deep to reach the original model they're working off of. Which turns out to be a conservative application of Linear No Threshold. Which... isn't actually testable for any reasonable value of statistical significance over the populations they're attempting to apply it to.

The BEIR VII risk models are a combination of excess relative risk (ERR) and excess absolute risk (EAR) models, both of which are written as a linear function of dose, depending on sex, age at exposure and attained age. The BEIR VII risk models were derived from analyses of data on the Japanese atomic bomb survivors for all cancer sites except breast and thyroid; for the latter, they were based on published combined analyses of data on the atomic bomb survivors and medically exposed cohorts.40, 41 To estimate risks from exposure at low doses and dose rates, a dose and dose-rate effectiveness factor (DDREF) of 1.5 was used for all outcomes except leukemia.

The biological effects of acute radiation exposure >1 Gy are reasonably well-known, are the basis for the linear-no-threshold model, and completely inapplicable to this sitation, as even the most-exposed workers at the Fukushima accident site did even approach this dose, despite the multiple situations where workers were exposed to doses in excess of legal limits.

The biological effects of short term dose less than 0.05 Gy or low-dose long-term exposure are also reasonably well-known, in that there is no statistically significant effect.

Unless you're dealing with the aftermath of a global thermonuclear war, the linear-no-threshold model is nearly useless from an epidemiological perspective, and so are conclusions reached using it.

Comment: Re:speed of light (Score 1) 374

by Behrooz (#46352425) Attached to: Report: Space Elevators Are Feasible

Technically, 29.9792458N Latitude constitutes a ~1cm band around the northern hemisphere. The Grand Gallery is oriented roughly north-south, and at 46m long in itself occludes 0.00043 latitude-- so you'll miss almost all of it.

Fortunately, there is a much more useful application for random decimal numbers associated with SI constants. If you happen to be flying over Africa and become lost, follow your GPS to the scientific notation of the Planck constant degrees east, then fly north, and you'll eventually reach Mohamed Boudiaf International Airport in Algeria.

Comment: Fuck the 'rule of law.' (Score 1) 822

by Behrooz (#46086757) Attached to: Ask Slashdot: What Does Edward Snowden Deserve?

I think you have a misunderstanding. In our society, the 'rule of law' is based more on color, wealth, politics, connections, and whether the justice system 'likes' you than on innocence or guilt.

*fuck* the rule of law. Laws have no judgement, rationality, and are subject to incredibly selective enforcement by cops, prosecutors, and the coercive apparatus of the government as a whole, entirely at the whim of officials who have no accountability or responsibility.

You really want a system where innocence or guilt is decided based on social class and race? Because that's where we're at right now.

If you'd like to talk about the rule of law, we can talk about our broken court system, where innocent until presumed guilty is a legal fiction, and better than 90% of crimes are resolved with a plea-bargain rather than a trial.

It's all a sham, and 'justice' is entirely illusory unless you're wealthy, educated, and connected enough to game the system. This is one of the (relatively few) places where the tea party and libertarians really have a solid point.

Comment: Re:Arrested . . . but will he be charged? (Score 1) 670

by Behrooz (#45522783) Attached to: Driver Arrested In Ohio For Secret Car Compartment Full of Nothing

Unless there is dope residue in the car, there is no way that any prosecutor would ever charge this because there is no way they could prove the intent element.

Because the opportunity cost for the prosecutor to file charges is so high, and innocent people never plea-bargain to avoid the threat of trumped-up charges that could put them in jail for the rest of their life if the trial goes badly because their overworked public defender is unable to mount a successful defense.

Prosecutors will happily charge anything that they think will work as leverage for a plea-bargain. That's their job, that's how the system works, and that's how all of the incentives are set up.

Comment: No, legal fees are not 'taken care of if you win'. (Score 4, Insightful) 670

by Behrooz (#45522757) Attached to: Driver Arrested In Ohio For Secret Car Compartment Full of Nothing

HAHAHAHAHAHAHAHAHAHAHAH.

No, legal fees are not 'taken care of if you win'. The system doesn't care if you are innocent, the system cares about the system, and obviously anyone who ends up in court has to be a scumbag, right?

  For criminal defense cases, you may choose to be represented without charge by an overworked, underfunded public defender who has every interest in resolving your case as quickly as possible via plea-bargaining... regardless of guilt or innocence.

Or you may hire an attorney who is actually being paid to represent your interests, where the cheapest option available is typically in excess of a thousand dollars, substantially more for serious charges or if the case actually goes to a jury trial.

The vast majority of defendants in the American legal system do not have the financial resources to hire an attorney, which is why the vast majority of all criminal charges are settled by plea bargain. Prosecutors have every incentive to pile on the threat of every imaginable charge and use the uncertainty of the outcome of a trial as leverage to coerce a plea bargain, guilty or not, because it works, and because they are almost never held responsible for their unethical conduct even when they commit egregious acts like concealing evidence that would exonerate the accused.

Add in unconscionable levels of police malfeasance and corruption on nearly every level, and the result is a criminal justice system that is anything but just. Unless you've got plenty of money. Which is kind of the point.

Comment: Re:Hey California, I have a solution for you (Score 1) 752

by Behrooz (#45408047) Attached to: Sweden Is Closing Many Prisons Due to Lack of Prisoners

The per-capita homicide rate in NYC is actually lower than in any other major city in the USA, barely above the national average.

Chicago roughly comparable to Atlanta... and even Detroit is still safer than New Orleans. You can argue the reasons, but you can't argue that the deep south has more than its share of social problems.

Comment: Re:At what speed? (Score 1) 722

by Behrooz (#45246687) Attached to: Google: Our Robot Cars Are Better Drivers Than You

Human drivers are bad at following safely because their reaction times are:

- Subjectively difficult to estimate accurately.
- Wildly variant from moment to moment dependent on the situation.

People are bad at realizing how fast they can respond. People are easily distracted, and are not capable of continuous full attention to more than a small fraction of their visual field even under ideal conditions. If you're looking at the radio to change the station, you are not physically capable of perceiving many changes in the peripheral view you have of the road, no matter how your brain fools you into thinking "Oh, I can still see everything", when your actual reaction time has just jumped by an order of magnitude.

Robotic reaction times are easy to measure objectively, and are situationally invariant. The the only relevant factors in following distance are the expected stopping distance at speed in the current conditions, and avoiding situations where the vehicle could potentially be unable to avoid a collision if whatever it is following stopped at maximum decelleration. This is EASY compared to most of the problems involved in navigating an unpredictable and changeable landscape.

I'd be much happier with a robot car following me than any human driver, the Stig included. It doesn't take a lot of distance to be safe with 50ms reaction time and rangefinders that are capable of discerning relative acceleration on a millisecond basis to form decisions with.

If entropy is increasing, where is it coming from?

Working...