Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
shabble
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
50 ms
·
421.
▲
by
shabble
13y ago
See also: Satellite imaging test patterns[1] [1] http://www.clui.org/newsletter/winter-2013/photo-calibration...
422.
▲
by
shabble
13y ago
Ooops, I didn't mean to imply that MEMS/stirling engines would eliminate the gradient problem, it was something of a after-thought that they would be more efficient than the Seebeck junction.
423.
▲
by
shabble
13y ago
The cable isn't likely to be the limiting feature here. The method of energy conversion doesn't just require heat, it requires a temperature gradient across the device. Here they're putting the tip in the fire to heat it, a
424.
▲
by
shabble
13y ago
<pedant> Strictly speaking it's not the Peltier effect, but the Seebeck effect. One is the inverse of the other; Peltier junctions produce a temperature gradient when current is passed through, and Seebeck ones generate
425.
▲
by
shabble
13y ago
I believe they mean that if (foo = 0) ... generates a warning like this: warn.c:2: warning: suggest parentheses around assignment used as truth value which you should be paying attention to, in favour of yoda-comparisons i
426.
▲
by
shabble
13y ago
> As a mechanical engineer I see tons of mechanical components in the human body. Diaphragm pumps, ball and socket joints, pinned joints, tensile supports. The chemical engineer points at the vast complexity of enzyme catalysed reactio
427.
▲
by
shabble
13y ago
Friction pads would probably have much more predictable wear (potentially in excess of the expected lifespan of the critter), whereas the discrete, stress-concentrating nature of gears would put them at higher risk of a single weak spot lea
428.
▲
by
shabble
13y ago
The ol' DIY Data-Diode[1] I've heard of using serial lines/modems with the appropriate tx->rx cut, but I don't know if it would actually work for ethernet (maybe 10BaseT only?) [1] https://en.wikipedia.org&
429.
▲
by
shabble
13y ago
I came across a paper on the first GPU keylogger[1] recently, which is quite clever, but maybe not sneaky enough. Sneaky-wise, hypervisor/system management processor/other embedded stuff (SMBus, BIOS/EFI, etc) are persistent,
430.
▲
by
shabble
13y ago
My microcode experience is some years ago, and nothing nearly as complicated as modern Intel stuff, but depending on just what you can control, you might be able to induce excessive thermal gradients by working some bits flat out and disabl
431.
▲
by
shabble
13y ago
If you have some crappy early-generation CFLs that have been used a lot, they can take 10-15 mins to come up to maximum intensity (although that max is still pretty puny, output wise).
432.
▲
by
shabble
13y ago
I imagine there could be a similar reputation market along the lines of the write-my-college-paper services, in which you sell the bug to someone who can then take credit for it. Certainly I'm sure you could find some people who'd
433.
▲
by
shabble
13y ago
I wonder if it would be practical couple the CPU directly to an externally accessible thermal pad, and have some sort of docking station which includes additional cooling. There are probably better cooling options available (liquid, heatpip
434.
▲
by
shabble
13y ago
I've been told before that `od' is the most portable, although I prefer xxd in virtually all situations.
435.
▲
by
shabble
13y ago
it's quite a common pattern in commandline apps to check if the stdout is a tty or not, and do things differently if so. So when you pipe it through `cat', it does the no-fancy-included output, much like the colouring that `ls
436.
▲
by
shabble
13y ago
Depends how badly they want it. Use a 0-day remote exploit to get a login, exfiltrate data. Use a known, unpatched exploit to do same. Use 0-day/known local exploit with sophisticated spear-phish trojan to do same. Use leverage over do
437.
▲
by
shabble
13y ago
The problem with "assume the maximum worst everything" case is that it places constraints on you that are essentially certain to be overly broad. Building a workable threat model & defending against it isn't practical (or
438.
▲
by
shabble
13y ago
The better solution would (IMO) to have been to out-Top-Gear Top-Gear in some fashion, and use that exposure to also cover your point of contention. I think a suitable segment to parody would be the (rocket assisted) Mini Cooper Ski-Jump[1]
439.
▲
by
shabble
13y ago
Every time I'm forced to use a forum for something, I wish usenet hadn't succumbed to spam & binary warez. Many of the clients from 20 years ago have more (useful) features than modern forums. It's probably about time to
440.
▲
by
shabble
13y ago
Although originally envisioned in a naval-defence role, I could see it serving as a powerful weapon indeed in the hands of the buccaneers of The Crimson Permanent Assurance [1] [1] https://en.wikipedia.org/wiki/The_Cri
441.
▲
by
shabble
13y ago
If I remember rightly, this also happened in some regional releases of the film Trainspotting. For the opposite problem, try reading Iain Banks' Feersum Endjinn [1] [1] https://en.wikipedia.org/wiki/Feersum_Endjin
442.
▲
by
shabble
13y ago
I came across a long writeup on the murder conviction of Michael Morton[1] which relied heavily on withholding evidence from the defence, and lying about it under oath. Upon his exoneration (after serving ~25 years), an inquiry was held, an
443.
▲
by
shabble
13y ago
I think http://www.jammed.com/~jwa/tempest.html might be the same FAQ/info page (right author, anyway), although last updated in '99, and no obvious links to any code (except one link about sniffing CPU emiss
444.
▲
by
shabble
13y ago
Having thought about it a little bit more, I remember there are actually a few things being done about it. Certificate Pinning[1] (bundle your cert with Chrome/$browser) HSTS[2] (cache the cert you receive on this connection for $num d
445.
▲
by
shabble
13y ago
You may be thinking of the Ellis/Cocks/Williamson[1] invention of the RSA encryption and Diffie-Hellman key-exchange algorithms several years prior to their open publication. [1] https://en.wikipedia.org/wiki/
446.
▲
by
shabble
13y ago
The problem is partly that you personally aren't using just a widely adopted algorithm, you're using a specific implementation layered on some monster protocol stack with weird legacy support for "Look the other way and ROT
447.
▲
by
shabble
13y ago
The obvious solution is more technology! Neural Implants with TPM/DRM. Maybe we have it already, but your hippocampus doesn't think you have Need To Know. I know it's been written about, but I can't remember where.
448.
▲
by
shabble
13y ago
Works for me despite a global RefControl block, which is either a bug or feature :) (FF23+RefControl+NoScript+ABP)
449.
▲
by
shabble
13y ago
You might find "Rule 34"[1] has some interesting ideas for automation and control. [1] http://www.amazon.com/Rule-34-Charles-Stross/dp/0441020348/c...
450.
▲
by
shabble
13y ago
For even less fun, there have been numerous incidents of consumer accesspoints/routers hardcoding NTP servers by IP address[1][2]. [1] http://pages.cs.wisc.edu/~plonka/netgear-sntp/ [2] http://web.
More ›