derf_

Also according to, like, Irving Fisher in 1933: https://fraser.stlouisfed.org/docs/meltzer/fisdeb33.pdf

"I venture the opinion, subject to correction on submission of future evidence, that, in the great booms and depressions, each of the above-named factors has played a subordinate rôle as compared with two dominant factors, namely over-indebtedness to start with and deflation following soon after; also that where any of the other factors do become conspicuous, they are often merely effects or symptoms of these two. In short, the big bad actors are debt disturbances and price-level disturbances."

Here's my question: What are the next ten words of your answer? Your taxes are too high? So are mine. Give me the next ten words. How are we going to do it? Give me ten after that, I'll drop out of the race right now. Every once in a while... every once in a while, there's a day with an absolute right and an absolute wrong, but those days almost always include body counts. Other than that, there aren't many unnuanced moments in leading a country that's way too big for ten words."

--President Josiah "Jed" Bartlet, The West Wing Season 4, Episode 5 (2002)

  static int foo_impl(int arg, int **resource1, int **resource2) {
    *resource1 = (int *)malloc(sizeof(**resource1));
    if (*resource1 == NULL) return EXIT_FAILURE;
    /* Do something with resource1 (omitted)... */
    *resource2 = (int *)malloc(sizeof(**resource2));
    if (*resource2 == NULL) return EXIT_FAILURE;
    /* Do something with resource1 and resource2 (e.g.,
       stick them in a global hash table as key/value pairs,
       or whatever, omitted)... */
    return EXIT_SUCCESS;
  }

  int foo(int arg) {
    int *resource1;
    int *resource2;
    int ret;
    resource1 = resource2 = NULL;
    ret = foo_impl(arg, &resource1, &resource2);
    if (ret != EXIT_SUCCESS) {
      free(resource2);
      free(resource1);
    }
    return ret;
  }

This difference means that the AND is still faster than the MOD when used on signed integers, if the compiler is not smart enough to prove that the input will never be negative.

For example, with gcc 4.7.3 the resulting assembly is

  movl %edi, %edx
  sarl $31, %edx
  shrl $24, %edx
  leal (%rdi, %rdx), %eax
  andl $255, %eax
  subl %edx, %eax

Innovation would have been using Opus (full disclosure: I'm one of the authors of Opus).

The only difference is we don't use rANS (which didn't exist when we started), we use a different arithmetic coding technique that does not require divisions, that also handles probabilities that don't sum to a power of two, so even the traditional frequency count style of models work fine.

Glad to see others are recognizing that this is a good idea.

a) everything is moving to mobile, and

b) every mobile OS ships its own browser which is not Firefox.

We know from Netscape how hard it is to stay relevant against that kind of pressure, and there isn't even a way to cry "Anti-trust!" because none of these players are monopolies. They just all behave the same way.

When I went to get directions a few weeks back and they told me that

a) I couldn't use Classic anymore, and

b) the new UI only had a drop-down for arrival time with half-hour increments, instead of letting me type a precise time (trains often arrive a couple minutes after the hour), and

c) they removed the option for getting public transit directions at all (that had been there and working in Classic for years),

I said three strikes and you're out and switched to Yahoo.

If you have a statute that you believe allows you to surveil the entire population, how many orders does that require? As opposed to ones that require some standard of suspicion for an individual or group.

I don't mean to imply there aren't other programs or theories of authority, but that this statistics tells you exactly nothing about the importance of a particular provision.

http://yosefk.com/c++fqa/exceptions.html#fqa-17.1

http://yosefk.com/c++fqa/ctors.html#fqa-10.17

http://yosefk.com/c++fqa/exceptions.html#fqa-17.3

The FQA helped my crystallize a lot of the reasons I hated C++. Even if you like C++, it's a good idea to understand the FQA's arguments, for the same reason people play devil's advocate.

My favorite poll is this one: http://www.pewinternet.org/2015/03/16/americans-privacy-stra...

Buried near the bottom is, "60% believe it is acceptable to monitor the communications of American leaders."

Raise your hand if you think it's good for a healthy democracy for the executive branch to monitor the private communications of congresspeople, senators, and federal judges?

[0] http://www.pewforum.org/2010/09/28/u-s-religious-knowledge-s...

[1] http://www.annenbergpublicpolicycenter.org/americans-know-su...

The marginal cost in terms of power may be very small (compared to a rocket), but the infrastructure costs are enormous, and as a result you're severely capacity-constrained.

Current link to Mars Odyssey from DSS 35 in Canberra is 14.22 kbps (roughly equivalent to a modem from 1991).

They may not understand what you're looking for in their answers. Did you tell them when you asked? It was obvious to you, but it may not be obvious to them.

Once I get a few of these answers, I try to ask questions that require more background explanation to respond to. This requires some thinking on your part. If you've seen a project like theirs before, you can ask them how they solved some of the problems that you ran into on your own project. "Did you consider X? What would happen if..." If you are actively engaged and thinking, people will respond to that, much more than if you ask very open-ended questions like, "Tell me about X."

The point is to get them to talk about something they have spent hundreds of hours thinking about. Something on which they are the experts. They should be able to go into detail and tell you what they found interesting. Both of those things are very informative.

On the contrary. Walmart's policy to date has been to simply close the store:

http://www.commondreams.org/headlines05/0210-13.htm

They got in trouble for that one, so now they at least pretend to have other excuses:

http://www.cbsnews.com/news/union-walmart-shut-5-stores-over...

Trying to know everything does not work. Targeted, traditional police investigation works. Money diverted from the latter to the former is wasted and reduces safety.

It was worth it to me to install Greasemonkey for this script alone.

The latter is exactly what causes the former to be wasted. If demand drops (or even merely does not rise as quickly as expected), then the price drops. If you were relying on high prices when you made your decision to invest more money to increase capacity, then what previously looked like a good idea may now be a bad one.

When your break-even price is $60 to $65/barrel and prices fall below that, your net return is negative. But the best part is you don't even get to stop producing. The costs for most oil sources must be paid largely up-front, and your marginal cost may be only $20 to $30/barrel. So when oil drops to $50/barrel, you can choose between continuing to produce at a loss of $10 to $15/barrel over the life of the well, or stop producing and suffer a much, much larger loss. That keeps supply high and prices low.

One oil source that does not have large up-front costs is fracking. Or, more exactly, fracking sites only last 18 months or so before production slows to a trickle, so you have a much smaller time horizon to recoup up-front costs. So the amount of fracking will very likely decline quickly. But viewed from an economy-wide standpoint, there was still plenty of malinvestment here: drilling equipment rental, trucking, rail, etc., require large capital outlays but carried high prices when oil prices and thus demand was high. As demand drops, they may not be worth the amount invested in them.

I remember a Maori exhibit at the Auckland War Memorial Museum. The gist was, they had these never-ending violence cycles based on perceived points of honor, but as you say they weren't a huge deal because their weapons weren't very good and most conflicts had few fatalities. Then they got guns from the Europeans and suddenly had an actual problem. At this point, some of the chiefs converted to Christianity for, according to the museum, the sole reason that it let them 'turn the other cheek' without losing honor (and thus their leadership position).

I remember only a small amount from the rest of the museum, but that bit was fascinating.

[0] http://www.defensivepatentlicense.org/

That does no one any good except the seller. For any seller, this kind of thing is table-stakes (no one wants to sell someone a gun pointed at their head).

But if the seller is specifically concerned about their patent being used offensively on others (which I suspect was the original commenter's concern), then this provision is useless.

To be told I cannot go through security until a few moments before the flight departs (because it is not even open, or the gate not even assigned) means that anything that does go wrong will cause me to miss my flight.

Not to mention the expense of having screening apparatus at every gate seems a little inefficient. It means everyone on the flight tries to cram through the one line at the same time, whereas if it were in advance, it could be spread out in time and over more lines.

Windows: https://bugzilla.mozilla.org/show_bug.cgi?id=922314

Android: https://bugzilla.mozilla.org/show_bug.cgi?id=986381

Linux supports HW decoding via gstreamer, if you have gstreamer set up to use your hardware (this can take some work).

Table VI. Encoding Run Times for Equal PSNR_{YUV}, HEVC vs. VP9 (in %): 735.2

I.e., HM is over 7x slower than VP9's slowest settings. VP9 speeds have improved dramatically since then, and it is now being used real-time (Google is working on adding it to WebRTC), while still demonstrating significant quality improvements over prior generations.

See also https://news.ycombinator.com/item?id=9329580 w.r.t. the quality aspect.

The H.264 prediction loop includes some inherent rounding biases and noise injection. Using 10-bit instead of 8-bit reduces the injected noise by 12 dB, which makes a noticeable difference in quality-per-(encoded)-bit, even when the source and display are both 8-bit. I spent some time talking with Google engineers early on in the VP9 design process, and they commented that they had done some experiments, but did not see similar gains by using higher bit depths with VP9. I don't know if that's still true with the final VP9 design, though.

Both of those are much older encoders, of course (nearly 2 years old in the case of the PCS paper), and some of the settings they used were questionable, e.g., what HEVC calls "constant QP" actually varies the quantizer based on the frame type, while VP9 really uses the same quantizer, which can make a big difference on metrics. Talking to the authors at PCS, they re-ran results later with more relaxed QP settings, and VP9 got somewhat better, but still didn't catch up with HEVC (on metrics).

Keep in mind all of these results are from people who have spent significant time working in MPEG/ITU/JVT/JCTVC, and may have some inherent biases. Google's own results looked much better: http://ieeexplore.ieee.org/iel7/6729645/6737661/06737765.pdf... (sorry for the paywall, I'm not aware of a free version available online, tl/dr: 30.38% better rate than H.264, 2.49% worse than HEVC), but obviously come with their own set of biases.

I don't know how to explain the discrepancies in the two sets of results, but they at least demonstrate the magnitude of the differences you can obtain by varying how you do the testing.

At worst the federal government decides the line the advertising industry has been feeding them about being able to regulate themselves is just that, a line, and decides perhaps consumers need some protections backed by force of law.

That's the only thing that ever had a chance of making advertisers take DNT seriously.