Forgot your password?
typodupeerror

Comment: Re:brute force (Score 1) 306

by dhTardis (#44027943) Attached to: Google Aims To Cull Child Porn By Algorithm, Not Human Review
Let's see... even Wikipedia's example of a poor JPEG is 1523 bytes, so (accounting for metadata) at least 2^10000 (10^3000) possible images. Divide by whatever images/second you like (a billion? a billion billion?) and it's still "more universe lifetimes than you can imagine". "Take time" and "impossible" are not, in this case, mutually exclusive.

Comment: Driving While Epileptic (Score 1) 1176

by dhTardis (#42919181) Attached to: Driver Trapped In Speeding Car At 125 Mph
I don't know about France particularly, but some jurisdictions issue licenses in some such cases because the epilepsy is either controlled by medicine, otherwise irrelevant to driving (e.g., seizures only while asleep), or simply so rare (epilepsy simply means "has had more than one seizure, ever") that it would be overboard to deny a license. Extra restrictions (e.g., reduced term of license) and/or requirements (e.g., medical supervision) are common, as might be expected.

Comment: Re:My two cents... (Score 2) 518

by dhTardis (#42051687) Attached to: Climate Contrarians Seek Leadership of House Science Committee

People here keep SAYING that Latour is wrong... but not one of them -- not even one, and not you -- has even attempted to show how he is actually wrong. And until they do, I will continue to accept what appears to be very solid and legitimate math and science.

OK, here goes. Latour starts by complaining that Spencer's example fails to specify whether the cold surround is kept there by constant heat removal or by a thermostat. Of course, he then goes on to state (correctly) that it makes no difference: the heat input (from the electricity) is constant, therefore (in steady state, which is the only thing under consideration) the heat removed is the same constant regardless of any unpowered additions to the chamber. So the heat input to the wall must be equal in the one- and two-plate cases as well; in both cases its temperature is constant, and the heat removed is the same.

That heat is of course exactly the radiation from one or both plates. In the two-plate case, some of the chamber wall can see only the added plate, which I suppose everyone agrees cannot reach the same temperature as the heated plate. That portion of the wall therefore receives less radiation. As the total must be unchanged from the one-plate case, some other part of the wall must be receiving more radiation. We could suppose this to be that part of the wall that can see both plates edge-on. But, either by making the plates thin and close together, or by (as Spencer suggests to exaggerate the effect) making the added plate actually be a (partial) shroud around the heated plate, we can reduce that effect to irrelevance. Therefore the additional radiation received elsewhere on the chamber wall must come from an increased temperature of the heated plate.

Comment: Re:I Agree With This Law (Score 1) 1155

by dhTardis (#33801170) Attached to: British Teen Jailed Over Encryption Password

I don't see this a "self-incrimination" issue, after all DNA and biological samples can be taken against your will and you cannot refuse to provide it if its called for.

They can collect your DNA, but you're not required to tell them if you're a chimera. There's a difference between being the subject of an investigation against your will (which goes for your person and your effects), and being compelled to assist in it actively.

In exactly the same way, they can read your encrypted hard disk (with a warrant), and they can break your safe (with a warrant). In the latter case, they'll likely ask you to open it for them for the simple reason that you'd rather have a functional safe afterwards and they'd rather do less work (so everybody wins). However, this law differs by saying that if the cops can't break your safe, you have to help.

Comment: Re:SSL (Score 1) 185

by dhTardis (#32676554) Attached to: Coming Soon, Web Ads Tailored To Your Zip+4

Of course HTTPS with any certificate whatsoever is >= HTTP. My point was that by the time ISPs are willing to add headers to your HTTP requests, they already have no problem with tampering with your connection and so might not be deterred by encryption-only HTTPS.

I like your DNS idea, as far as protecting those sites that have CA-certs goes, but it seems a bit circular to me; in order to allow my browser to stop bothering me about self-signed certificates, I go find some CA certificates (which DNSSEC must use if my ISP is to be powerless against it) and use them to double-check the self-signed ones (to make sure that the site in question is expected to present one). Why not just go for the CAs in the first place?

Unfortunately, I bet that going to HTTPS would be a benefit. What I mean by that is that the law is (probably) only strong enough in the case of encryption, and doesn't do what it needs to: unequivocally forbid tampering with traffic that isn't yours (via Comcast's RSTs or these HTTP headers or Phorm or whatever).

(Incidentally, I wonder if there's a technical solution to this that one can implement on the assumption that the ISPs are being cheap about implementing it: deliberately fragment your packets (at TCP or IP level) and keep them from identifying (the assumption: from one packet at a time) a reliable place at which to insert their headers.)

Comment: Re:No (Score 1) 185

by dhTardis (#32664556) Attached to: Coming Soon, Web Ads Tailored To Your Zip+4

so less need to blast people in the face in order to get their attention.

But if blasting people in the face improves returns now (which it must, since advertisers keep doing it), why would it stop improving returns later? The only that that would help is if people paid enough attention even to subtle ads so that the annoyance of the blasting would outweigh the added eyeballs. Are ZIP+4s going to do that?

Comment: Re:SSL (Score 1) 185

by dhTardis (#32664536) Attached to: Coming Soon, Web Ads Tailored To Your Zip+4
I can't tell if you're being ironic or not. But if you are serious, then know that an ISP that's willing to stoop to this level would have a field day with this: they just have to intercept your connection, present you a self-signed certificate for whatever domain you tried to contact, and then they can do whatever they want (among which the least-harmful might be adding your ZIP+4 to the headers). It is malfeasance like this that the PKI is designed to prevent! (Separate topic: how trustworthy are the CAs?)
Patents

+ - Intellectual Ventures, King among Patent Trolls->

Submitted by
Nrbelex
Nrbelex writes "The New York Times is running a profile of Intellectual Ventures, the largest company within the category of firms that hold patents, but do not make products. Nathan Myhrvold, leader of Microsoft's technology development in the 1990s and current chief of the company claims the patent world is a vastly underdeveloped market, starved for private capital and too dependent on federal financing for universities and government agencies, which is mainly aimed at scientific discovery anyway. Eventually, he foresees patents being valued as a separate asset class, like real estate or securities. Yet while Mr. Myhrvold is saying one thing, his company’s main activity is quite another, according to Mark Bohannon, general counsel and senior vice president for public policy for the Software and Information Industry Association."
Link to Original Source
Privacy

PA School Spied On Students Via School-Issued Laptop Webcams 941

Posted by timothy
from the wait-for-your-health-insurance-computer dept.
jargon82 writes "A Pennsylvania high school is using laptops they issued to students to spy on them in homes and outside of school. According to a class action filling the webcams and microphones in these laptops could be remotely activated by school officials, and have been used in this role. One student was accused of 'improper behavior in his home' and the school provided a photo taken via his laptop as proof."

Comment: Re:Kind of One Sided Review of the Service (Score 1) 495

by dhTardis (#30769950) Attached to: Tynt Insight Is Watching You Cut and Paste

1) you are copying and pasting Wired's content and 2) as early as high school I was taught that if I was copying information verbatim, I had better have some sort of reference

  • What if I see a phrase on Wired's site and decide to search for it on another site?
  • What if I'm using someone else's machine, and the best I can do to defeat a potential keylogger is grab random letters from random places?
  • What if I'm bookmarking their page and want some useful text for the description?
  • What if I'm writing a point-by-point rebuttal and don't need to cite them after every quotation?
  • What if I'm selecting a URL that isn't a hyperlink so that I can direct my browser there?
  • What if I'm copying a quotation in their copy, and would prefer to cite the original source?
  • What if I'm writing an email to a friend recommending the piece, and would like to include my favorite part as a hook?
  • What if the article isn't written in my native language and I want to use a translator program on certain words?
  • What if I want to copy a command into my terminal? (This is Wired, after all.)
  • What if I'm doing statistics on journalistic writing and am feeding their prose into a calculator of the Gunning fog index? (Should I have to include a link to Wikipedia there because it's where I retrieved the name of the index? They don't own it.)

Provide an easy-to-use "cite us" link if you like, to encourage proper citation practice. But it is the height of narrow-mindedness to assume that you know what other people want to do with text you provide them, and of arrogance to assume that you know how to do it better than they do.

Comment: Derivative works (Score 2, Informative) 371

by dhTardis (#30643612) Attached to: Monty Wants To Save MySQL

Personally, I feel the most reasonable interpretation (from the standpoint of being consistent and, well, logical) would be that linking does *not* create a derivative work (for example, is Firefox a derivative of the Flash plugin, or Flash plugin a derivative of Firefox? Seems to me they are fairly independent works that use the mechanism of dynamic linking to work together.)

No one claims that Firefox is a derivative of Flash, or vice versa. What is claimed is that the resulting memory space with both objects loaded is a derivative work of each, which can only be created with the license-granted consent of the copyright holders of both objects. The trick: how strong is this argument since it applies only to the ephemeral address space created at runtime by the user and not, say, by Mozilla or Adobe?

In the obvious case of MegaCo distributes foo that always links dynamically to libgpled.so, one can argue that MegaCo is effectively creating that combined object because there is no other way that their software could be used. But IANAL, of course, and I think the real lawyers (and judges) haven't fully settled the question.

Comment: Re:Passive propulsion (Score 1) 322

by dhTardis (#29702947) Attached to: Hyperdrive Propulsion Could Be Tested At the LHC

Why does the second law say no to arbitrarily concentrating energy?

Because it lets me extract useful work simply from a hot environment with no cold reservoir. I just put my box in a room and it gradually gets hotter while the room gets colder. After a while I set up a heat engine between them and extract some work from the difference. Then I can repeat the process: this time, the room can't heat the box as well, but the room is colder too, so the heat engine still works. Eventually I can get (almost -- see the 3rd law) all the heat from the room extracted as useful work, which is not allowed.

(By "arbitrarily" I meant "without reason or effort" rather than merely "without bound". Sorry for any confusion.)

Surely the fact that you are drawing all your energy from outside the system (your box) would allow for a localised loss of entropy, given the gains in entropy outside the system.

Entropy goes the other way: any normal physical system that gets hotter gains entropy, so the box would be gaining it and the room losing it. But hot systems (by definition!) gain less entropy from, say, a joule of heat than cold systems do. This is the reason for the unidirectional flow of heat: if I'm hotter than you, you lose more entropy giving up a joule than I gain from getting it, so we can't send the heat that way.

"Kill the Wabbit, Kill the Wabbit, Kill the Wabbit!" -- Looney Tunes, "What's Opera Doc?" (1957, Chuck Jones)

Working...