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

 



Forgot your password?
typodupeerror
×

Comment Re:That's not how science works (Score 2) 141

Nothing has been proven.

Correct.

Science is the discipline of publicly testing ideas by systematic observation, controlled experiment and Bayesian inference. As such, proof is simply not relevant to what it does, which is produce knowledge. Knowledge--unlike faith--is inherently uncertain.

It'll take a few hundred years for the popular science press to catch up to this. What is being presented here is evidence that the idea p-p fusion powers the sun is correct, so the posterior pluasibility of that idea goes up, although not to 1 (which would be a certainty, and therefore an error: an idea that was immune to additional evidence.)

If neutrinos had not been detected, the plausibility would have gone down, although not to 0 because that would be the same error. Science never disproves anything any more than it proves anything. Proof and certainty are like the philosopher's stone sought by alchemists: a fundamentally mistaken goal.

Philosophers are the alchemists of epistemology, discovering all kinds of cool things while on a hiding to no-where.

Comment Re:And this is how we get to the more concrete har (Score 1) 528

I think you'll find that the utility of falsification is established philosophically, not by observational fiat.

But falsification is at best marginally relevant to science, which is the discipline (not method) of publicly testing ideas by systematic observation, controlled experiment and Bayesian inference.

Falsification simply never comes into it, outside of outlandishly models of science promoted by ignorant philosophers. Promoters of "scientific method" and falsification are almost never scientists, and most scientists will quietly ridicule the ideas if you give them a couple of beer.

Bayesian logic is established by mathematical deduction using an argument from invariance of a kind that originated within mathematical physics (Einstein's arguments for relatively are the most famous of the kind).

Philosophers don't even have the right goal--they are always running after "certainty", which we now know to be a epistemic error. Knowledge is not certain and cannot be certain, because only Bayesian reasoning can produce knowledge, and Bayesian reasoning is not capable of producing a posterior plausibility of 1 or 0 (ie certainty).

Comment Re:Statistics as standalone field (Score 1) 115

If your view that "everyone is complete shit at statistics", that should include statisticians.

This has been my experience as well. I would go so far as to say that statisticians understand probability less well than most working experimental scientists. They are overly-enamoured of abstract models and rarely dig down to the raw probability distributions underneath, which is what working scientists actually care about.

AT&T

Ask Slashdot: What To Do About Repeated Internet Overbilling? 355

An anonymous reader writes "AT&T has been overbilling my account based on overcounting DSL internet usage (they charge in 50 gigabyte units after the first 150). I have been using a Buffalo NFinity Airstation as a managed switch to count all traffic. As you may know, this device runs firmware based on dd-wrt and has hidden telnet functionality, so I am able to load a script to count traffic directly onto the device. I have an auto-scraper that collects the data and saves it on my computer's hard disk every two minutes while the computer is running. While it is not running, the 2 minute counters accumulate in RAM on the device. Power problems are not normally an issue here; and even when they are I can tell it has happened. The upshot of all this is I can measure the exact amount of download bandwidth and a guaranteed overestimate of upload bandwidth in bytes reliably. I have tested this by transferring known amounts of data and can account for every byte counted, including ethernet frame headers. AT&T's billing reporting reports usage by day only, lags two days, and uses some time basis other than midnight. It is also reading in my testing a fairly consistent 14% higher whenever the basis doesn't disturb the test by using too much bandwidth too close to midnight.

AT&T has already refused to attempt to fix the billing meter, and asserts they have tested it and found it correct. Yet they refuse to provide a realtime readout of the counter that would make independent testing trivial. I've been through the agencies (CPUC, FCC, and Weights & Measures) and can't find one that is interested, AT&T will not provide any means for reasonable independent testing of the meter. It is my understanding that if there is a meter and its calibration cannot be checked, there is a violation of the law, yet I can't find an agency that can even accept such a claim (I'm not getting "your claim is meritless", but "we don't handle that"). If indeed they are not overbilling, my claim of no way to verify the meter still stands. My options are running thin here. So that my account can be identified by someone who recognizes the case: 7a6c74964fafd56c61e06abf6c820845cbcd4fc0 (bit commitment).

Comment Re:The death of leniency (Score -1) 643

The problem with this is that if all cops feel like they're being audited all of the time, they're less likely to let you off the hook for a minor violation.

That's not a problem. Selective enforcement, or IOW cops letting people off the hook for minor violations, leads to a lack of respect for the law, and the proliferation of bad laws which always end up used to punish some classes and not others on the basis of prejudice.

If rich mofos were prosecuted for bullshit crimes, you'd see those crimes fall off the books.

Comment Re: Not the PSUs? The actual cables? (Score 1) 137

And, once again, Martin demonstrates his learned discourse and debating skills.

You don't get to insult me and then bitch, piss and moan when I return the favor.

Are you really this much of an asshole, or is it just on Slashdot?

I usually only encounter people willing to act like as much of a lame as you have been in this conversation on slashdot, so it's pretty much just here.

Hardware

Video Slashdot Talks WIth IBM Power Systems GM Doug Balog (Video) 36

Yesterday we had a story titled 'IBM Gearing Up Mega Power 8 Servers For October Launch.' In the intro Timothy wrote, '...watch for a video interview with Balog on how he's helping spend the billion dollars that IBM pledged last year on open source development.' This is that video, and in it Balog tells us how much IBM loves Linux and open source, and how they're partnering with multiple distros, recently including Ubuntu. So get ready for Power 8 servers in October. IBM is pushing them like mad -- especially in the Linux/FOSS realm. (Alternate Video Link)

Comment Re:I like... (Score 1) 643

(including an officer recall if the feed fails)

No, including "you are only a uniformed officer while the feed is live. If it stops, you are a regular private citizen until it's restarted".

Then it's not police brutality. It's armed citizen brutality, without all the protections of the badge.

Comment Re:anyone remember when (Score 1) 316

My first computer with a hard drive was an Amiga 2000 that came with a 120MB Maxtor. I was gleeful at its blinding speed and unfathomable capacity compared to my older floppy-based system. So much so, in fact, that I spent quite a few hours brilliantly doing the AmigaDOS equivalent of cp -R /media/floppy / so that I'd never have to bother with those slow things again.

That was perhaps my first introduction to the importance of namespaces, a lesson which I carry with me unto this day.

Slashdot Top Deals

"Ninety percent of baseball is half mental." -- Yogi Berra

Working...