Hacker Newsnew | past | comments | ask | show | jobs | submit | more lambda's commentslogin

OK, that's freaky. I read the HN comments first, then the original article. By the time I read the original article, it had been updated to quote the HN thread:

  On Hacker News, the grumbling about the settlement was 
  immediate. “I honestly do not understand why the 
  plaintiffs would settle this case,” one poster wrote. But 
  some saw another side: “What did the engineers risk with 
  this lawsuit? Nothing. What did the law firm risk? 
  Getting paid peanuts for hundreds of hours they spend on 
  the case if they lose.”
I did not know that the NY Times had started updating their articles with quotes from discussions about the same articles. Do they base which quotes they choose on your referrer? Will I see different quotes if I view this linked from Reddit?


I read the article first, and thought they were referring to another older HN thread on this topic. Thanks for pointing this out. It's a bit hard to believe (and insane) that they're updating their article real-time. And they've not even marked it as an edit.


Also note how cavalierly and routinely the Times (and most other "traditional" publishers) avoid actually linking to this thread, or basically anything other than internal links to other Times articles.


Sure, it's not going to work for the majority of OSS developers; but nothing is. It's nice to have different kinds of opportunities available. Some people can do stuff in their free time; some may be more productive with a dedicated sprint. Some people may have the opportunity to do it between jobs; some developers are consultants, and can probably find a way to free their schedule up for a few months if they can justify it with the extra income.

Different companies contribute in different ways. The company I work for funds major features that we need, as well as having support contracts with the primary developers of some of the key software that we use, and we contribute back patches when we find bugs that need fixing. Google does their Summer of Code, in which students can work remotely over their summer break, plus does primary development of a lot of projects like Android and Chrome, and contributes to others like the Linux kernel. Red Hat and Canonical mostly develop open source software directly. The Gnome OPW funds internships for women, over several periods over the course of the year.

So there are lots of opportunities out there, with a lot of different ways to get funding and support. I don't think that each program needs to be all things for all people; it's OK to have particular targeted and focused programs, either on different groups of developers, different projects, different types of time commitment, and so on.


I completely agree with everything you said. I got the impression from the blog post they are looking for people to solve real world problems, and people who are good at that are generally employed (comfortably). I hope the program is successful and they expand it at some point to allow for different types of arrangements.


  In my opinion, text adventures can have a certain beauty 
  because they tell a story in simple words (for a great 
  example, see Photopia).
I never managed to get the appeal of Photopia. It just grated on me. Never could get the hang of Adam Cadre's style; always preferred Emily Short or Andrew Plotkin.


When I discovered it, I hadn't done much interactive fiction for a while, and for some reasons it really got to me. I had no idea what to expect going in, and the experience of exploring it, even if it's essentially non-interactive, was enlightening. I loved when it occurred to me to type "fly". It was liberating. Like all good art, the experience it gave me is essentially incommunicable.


> always preferred Emily Short or Andrew Plotkin.

Likewise, as well as Nick Montfort; Ad Verbum is exceptional.


What makes you say that this kind of study can never ascend to the level of science? This kind of study is science. It's imperfect, but all of science is; science is just a collection of tools to improve our knowledge of the world, coupled with the body of knowledge so gained. Some of those tools are stronger than others, or can provide more certainty.

Controlled studies in which you vary only one input to a system at a time, and observe the effect, are obviously a stronger tool than mere observational studies. But that doesn't make observational studies not science.

It's perfectly possible to determine which of these effects are more or less likely to be responsible; the different hypotheses that you make would make easily testable predictions, which can be tested independently because different jurisdictions have made these changes at different times. Lead was banned in different countries at vastly different times (https://en.wikipedia.org/wiki/Tetraethyllead). Likewise, three-strikes laws have been introduced at different times in different places, and many places have never introduced such laws (https://en.wikipedia.org/wiki/Three-strikes_law). Easy availability of birth control and abortion may be a little harder to quantify precisely, but that can be done too.

Since some countries have only banned leaded fuel in the past 10 years, within another 10 or 15 we should be able to say with considerably more certainty whether this has a causal relationship. If the crime data for countries which banned it later follow the same pattern, that's a very strong indication that the theory is correct.

I would find it hard to believe that the three strikes laws would have the same effect, as the reduction in crime has been happening fairly consistently nation-wide, while three strikes laws have been implemented at vastly different times in different states, or not at all. Lead levels are much more consistent nationwide, as those have mostly been national laws (with the exception of California banning leaded gasoline a few years earlier).

Anyhow, I strongly disagree that you can say that this isn't science. It's imperfect science; anything as complex as the social phenomenon of global crime levels is going to have a lot of complex, interacting factors and no one of them is going to be completely explanatory, and you are never going to be able to do significant studies with controlled inputs. But there are scientific tools for dealing with these somewhat more ambiguous situations, and we can and should apply them as the policy implications are quite substantial.


> What makes you say that this kind of study can never ascend to the level of science?

Because the uncontrolled nature of the observations means evidence cannot be reliably collected in a way that separates one cause from any number of unintended causes and side effects. The only way to get around these obstacles would be to feed one group of people doses of lead and compare them to another in a controlled setting, or isolate two groups of people, give one group the right to birth control and abortions, deny the other group those rights, and keep them from finding out about each other. Things that cannot be done for ethical reasons, and things that prevent social psychology from rising to the level of science.

This is why social psychology cannot be a science, and it explains why the field has the terrible reputation it has and deserves.

Link: http://chronicle.com/article/As-Dutch-Research-Scandal/12974...

Quote: "Even before the Stapel case broke, a flurry of articles had begun appearing this fall that pointed to supposed systemic flaws in the way psychologists handle data. But one methodological expert, Eric-Jan Wagenmakers, of the University of Amsterdam, added a sociological twist to the statistical debate: Psychology, he argued in a recent blog post and an interview, has become addicted to surprising, counterintuitive findings that catch the news media's eye, and that trend is warping the field."

> Anyhow, I strongly disagree that you can say that this isn't science.

Yes, I understand, but you haven't the slightest idea what makes science science. If you did, you would recognize this sort of article for what it is.

Social psychology deserves its reputation, science operates at a higher level, this is Science 101, and you have no idea what you're talking about. Not a clue.

There is a one-word summary of what science requires, that word is known to every qualified scientist, and you do not know what it is. How do I know this? You defended social psychology for over four hundred words without even mentioning it.


> Because the uncontrolled nature of the observations means evidence cannot be reliably collected in a way that separates one cause from any number of unintended causes and side effects. The only way to get around these obstacles would be to feed one group of people doses of lead and compare them to another in a controlled setting, or isolate two groups of people, give one group the right to birth control and abortions, deny the other group those rights, and keep them from finding out about each other.

Uh, no, statistical controls are a real tool of real science. While laboratory conditions are in some ways (ethics aside) ideal, they aren't a prerequisite for real science.


> Because the uncontrolled nature of the observations means evidence cannot be reliably collected in a way that separates one cause from any number of unintended causes and side effects.

If we couldn't scientifically deal with this scenario, we'd hardly do any science. The world is noisy, measures are inaccurate, life is not perfect, but we learn to deal with it. Not all science can be math theory. When it touches the real world, you have to deal with imperfection.


> If we couldn't scientifically deal with this scenario, we'd hardly do any science.

Feel free to change the topic. Science is very clearly defined, and an absence of falsifiability is fatal.

Either:

1. A theory can potentially be falsified by new evidence, or

2. IT IS NOT SCIENCE.

http://en.wikipedia.org/wiki/Falsifiability

Quote: "The concern with falsifiability gained attention by way of philosopher of science Karl Popper's scientific epistemology "falsificationism". Popper stresses the problem of demarcation—distinguishing the scientific from the unscientific—and makes falsifiability the demarcation criterion, such that what is unfalsifiable is classified as unscientific, and the practice of declaring an unfalsifiable theory to be scientifically true is pseudoscience."

Which part of the above didn't you understand?

> When it touches the real world, you have to deal with imperfection.

As I said, feel free to change the subject.


No need to be rude - you have come to an invalid conclusion here, and it would be better to try to understand why than yell at others.

So, what about this research is not falsifiable?

For example, if research finds another location in which lead additives were not banned, but in which crime declined with the same characteristics as the drop being ascribed to the ban on lead in other countries, that's strong evidence that the theory is indeed false.

In other words, your point is entirely incorrect.


> you have come to an invalid conclusion here ...

It's time to offer evidence, not opinion. My claim (and that of scientific philosophers) is that falsifiability is requires for science. That is not a point that's open to debate, it is an established requirement. It's even written into the law, for example laws that prevent Creationism from being presented as science in public school science classrooms:

http://en.wikipedia.org/wiki/McLean_v._Arkansas

Quote: "The judgment defined the essential characteristics of science as being:

   * It is guided by natural law;
   * It has to be explanatory by reference to natural law;
   * It is testable against the empirical world;
   * Its conclusions are tentative, i.e. are not necessarily the final word; and
   * It is falsifiable."
> So, what about this research is not falsifiable?

Try to imagine how someone could prove false the claim that a reduction of lead in gasoline caused a reduction in the crime rate, to the exclusion of other factors, in a social setting, with no meaningful controls in place.

> In other words, your point is entirely incorrect.

You need to make an argument before coming to a conclusion. At least, that's the rule in a scientific debate.


Try to imagine how someone could prove false the claim that a reduction of lead in gasoline caused a reduction in the crime rate, to the exclusion of other factors, in a social setting, with no meaningful controls in place.

I literally gave you an example of this already, and you've ignored it. I can only assume you're responding in bad faith.


>> Try to imagine how someone could prove false the claim that a reduction of lead in gasoline caused a reduction in the crime rate, to the exclusion of other factors, in a social setting, with no meaningful controls in place.

> I literally gave you an example of this already, and you've ignored it.

You're ignoring the requirement to "prove false". Your example doesn't meet this criterion, and the criterion is required for science. Here's your example:

"For example, if research finds another location in which lead additives were not banned, but in which crime declined with the same characteristics as the drop being ascribed to the ban on lead in other countries, that's strong evidence that the theory is indeed false."

No, that's not "strong evidence", it merely introduces another confounding factor, and it is not remotely a basis for falsification. It's not a black swan.


You're ignoring the requirement to "prove false"

Unless you are claiming that all observational science is, in fact, not science, then this idea is intractable.

For example - cosmology. We can observe the universe. We can't create any meaningful controls when we do so, but we can use the observed data to construct models. If those models fit our observations, then they are accepted. If we find a situation in which the model does not correctly describe a system (i.e. it has been falsified), then we can modify it, or discard it entirely.

One can adopt entirely the same process to research the effects of lead. We can observe that, based on available data, a certain phenomenon occurred. We can use other sources of information to compensate for other possible causes. If there is information available which contradicts that conclusion, then we can discard the model. Like other scientific disciplines, there is never 100% certainty about the validity of a model, however if adequate contradicting evidence can be shown, then it will be shown invalid.

The "black swan" is a simplistic view of falsifiability. When presented with this swan, one can respond "Ah, you are correct. This is a swan, and it is black; my model for swan coloration is invalid." One could also legitimately respond "Ah, you are not correct; this swan is white, and has been exposed to dye which has made it appear to be black." That is to say that falsification is never absolute, and all observations are subject to confounding factors.

Real science rarely falls into clean classification of certainty and uncertainty; there are always confounding factors, and the model of lead's effect on crime is not qualitatively different.


What makes you say the results of these studies are unfalsifiable?


Try to imagine how you would prove false the claim, in a social setting, with multiple factors at work, and no meaningful controls. If I shake a dried gourd over a cold sufferer and he gets better, I can claim to have been the cause of the recovery. And guess what? No one can prove that claim false. The fact the colds are really caused by germs is true and interesting, but it doesn't falsify my claim to have been the cause of the recovery. That's why scientific claims requires a basis for falsification.

And this is more than a philosophical position, it's written into the law:

http://en.wikipedia.org/wiki/McLean_v._Arkansas

Quote: "The judgment defined the essential characteristics of science as being:

   * It is guided by natural law;
   * It has to be explanatory by reference to natural law;
   * It is testable against the empirical world;
   * Its conclusions are tentative, i.e. are not necessarily the final word; and
   * It is falsifiable."


>If I shake a dried gourd over a cold sufferer and he gets better, I can claim to have been the cause of the recovery. And guess what? No one can prove that claim false.

Umm, sure they can? They repeat the experiment (or perform a better one) and don't get the same results. Lack of controls makes for weaker evidence, but it doesn't magically make things unfalsifiable.

And so I ask you again: What makes you say these studies are unfalsifiable?


> Umm, sure they can? They repeat the experiment (or perform a better one) and don't get the same results. Lack of controls makes for weaker evidence, but it doesn't magically make things unfalsifiable.

Yes, in fact, that's exactly what it does. My dried gourd example should have demonstrated why this is so -- I can always claim a correlation with the gourd and the outcome, and without proper experimental controls, that's irrefutable.

My dried gourd example is meant to show the need for experimental controls -- another group that recovers in the same amount of time without benefit of gourd-shaking undermines the claim that the gourd had anything to do with the outcome. But "undermines" doesn't equal "falsified" -- the latter condition requires that we test an explanation, not just an observation.

The method of performing more experiments, gathering more data, decreases the p-value of the outcome, but it cannot turn a hypothesis into a theory until we're testing an explanation rather than measuring an effect. As long as we measure effects without trying to explain them, we're doing an expensive kind of accounting, but not science. Science requires that we try to offer and then test an explanation for what's being observed, and toss the explanations that don't pan out.

The tl;dr: science requires theories.

> And so I ask you again: What makes you say these studies are unfalsifiable?

The problem we're having here involves a definition of terms -- what falsifiable means. More here: http://arachnoid.com/building_science


I thought this argument sounded familiar, and looking at my comment history confirms it. I argued with you about the necessity of an explanation for something to be scientific previously. Clearly you feel strongly about this issue, so I'll make a better attempt to understand your argument this time.

Can you clarify for me exactly what you mean by an "explanation"? The linked article starts to talk about a difference between description and explanation, but then just arbitrarily says that Newton's theory of gravity constitutes an explanation. I don't see how Newton's theory is anything more than a very detailed description.

Why is the hypothesis "There's a magic force called gravity that behaves according to these laws." any more valid than the hypothesis "Stuff falls because magic." Both provide testable predictions, and can be falsified. Indeed one could simply point to the moon and say "that doesn't fall, theory disproven" just as one could point to the unusual orbits, and disprove Newton. I don't see how that makes the hypothesis itself unscientific.


You could I guess define Science as only applying to areas where you can do controlled experiments - but you would thereby eliminate all observational sciences - evolutionary biology, astrophysics, cosmology, etc.

I suspect most people would find that a somewhat extreme position.

Also, fraud is an issue in all fields of Science, so that seems somewhat an unfair issue to introduce.


> but you would thereby eliminate all observational sciences - evolutionary biology, astrophysics, cosmology, etc.

No. Not if an observation leads to a testable, falsifiable theory, if that theory makes new predictions, if the predictions are verified, and if the theory in all its manifestations remains potentially falsifiable by new evidence. Like quantum theory, the best scientific theory we have, a theory that makes many other fields look like stamp collecting, and a theory that relies on observations of real-world phenomena.

As to cosmology, there are a number of cosmological theories based on observation that have led to testable theories, and those theories have survived reality-testing and have made predictions that are later verified in observations, like the recent observations of the CMB that support inflationary theory and led to some recent Nobel Prizes.

> I suspect most people would find that a somewhat extreme position.

Science is not a popularity contest, and it is not politics.

> Also, fraud is an issue in all fields of Science, so that seems somewhat an unfair issue to introduce.

Many people don't understand that some pseudoscientific fields are all fraud, in which case arguing that all fields have some fraud misses the point.


The hypothesis that lead exposure affects early childhood development, and that this leads to higher crime rates does lead to testable predictions - it predicts, for example, that different regions that removed lead at different times would see crime rates drop at different times. And indeed, this is exactly what we see. Is this as strong as exposing a test population and comparing it to a control? No. But neither is it meaningless pseudo-science.

This line of argument is actually very similar to the cosmological example - the theory makes a prediction about something you should observe, if the theory is correct. However, this is still much weaker proof than being able to generate a Big Bang in the lab and examine the outcome.

The QM case is an example of how much more powerful it is in fields where we can do controlled experiments, I agree. But not everything worth investigating (and worth knowing) is amenable to that type of study.


> The hypothesis that lead exposure affects early childhood development, and that this leads to higher crime rates does lead to testable predictions ...

Yes, but not falsifiability, required for science. Okay, I see people aren't getting this, so here's a classic example. Let's say I am a doctor who believes he has found a cure for the common cold. My cure is to shake a dried gourd over the cold sufferer until he gets better. The cure might take a week, but it always works. My method is repeatable and perfectly reliable, and I've published my cure in a refereed scientific journal (there are now any number of phony refereed scientific journals). And, because (in this thought experiment) science can get along without defining theories, without falsifiability, I'm under no obligation to try to explain my cure, or consider alternative explanations for my breakthrough — I only have to describe it, just as lead in gasoline is being described as inversely correlated with crime rates in the linked article.

> Is this as strong as exposing a test population and comparing it to a control? No. But neither is it meaningless pseudo-science.

But that is exactly what it is -- meaningless pseudoscience -- on the ground that no one is crafting and then testing theories about the correlation. Until we know why A is correlated with B, until we can demonstrate a cause-effect relationship and make falsifiable predictions based on the theory, it's not science.

> But not everything worth investigating (and worth knowing) is amenable to that type of study.

Yes, I agree, science can't provide answers to questions not amenable to the scientific method (i.e. "that type of study"), and not all important questions have this property.


Well in the case of lead wouldn't the prediction be: a jurisdiction that removes lead from its petroleum products will experience a significant decrease in crime rates about 20 years later?


Not without a testable, falsifiable explanation. Observations (i.e. descriptions) can't be falsified, that's reserved for explanations. If I say, "the night sky is filled with little points of light", that's hardly falsifiable. But if I say, "those points of light are actually thermonuclear furnaces like our sun, but at greater distances", that claim is open to meaningful test and possible falsification.

Here's why a testable, falsifiable theory is required for science. Let's say I'm a doctor and I've created a revolutionary cure for the common cold. My cure is to shake a dried gourd over the cold sufferer until he gets better. The cure might take a week, but it always works. My method is repeatable and perfectly reliable, and I've published my cure in a refereed scientific journal (there are now any number of phony refereed scientific journals). And, because (in this thought experiment) science can get along without defining, falsifiable theories, I'm under no obligation to try to explain my cure, or consider alternative explanations for my breakthrough — I only have to describe it, just as the linked article describes the correlation between lead in gasoline and the crime rate.

Because I've cured the common cold, and because I've met all the requirements that social psychology recognizes for science, I deserve a Nobel Prize. Yes or no?


I welcome your strong defence of best quality science. The general public needs to know how the stronger sciences differ from the softer sciences.

However, this post is needlessly rude.


  The only way to get around these obstacles would be to 
  feed one group of people doses of lead and compare them to 
  another in a controlled setting, or isolate two groups of 
  people, give one group the right to birth control and 
  abortions, deny the other group those rights, and keep 
  them from finding out about each other. Things that cannot 
  be done for ethical reasons, and things that prevent 
  social psychology from rising to the level of science.
And yet in another comment, you say that cosmology, which has the same problem with not being able to vary the inputs and observe the effects on the outputs, is a real science. Why? Because it offers falsifiable hypotheses; you can make a prediction based on certain data, then test that prediction against other data.

The exact same is true of this topic. As I pointed out in my comment above, in portions that you conveniently ignored, there are strong predictions that this research makes, which can be tested. The original paper on this topic was written in 1999, and there is already 15 years worth of additional data since then; the next 10 to 15 years will provide even more data, which could falsify or could be consistent with this hypothesis.

  This is why social psychology cannot be a science, and it 
  explains why the field has the terrible reputation it has 
  and deserves.
I don't know why you bring up social psychology, as this study is not social psychology. It's more environmental sociology. Social psychology is the study of how people's behavior is related to their beliefs and attitudes about other people. This study is about how chronic, population wide lead poisoning relates to criminal behavior.

And while I share your skepticism about social psychology, the mere presence of corrupting influences such as the ones you mention, the desire for snappy, headline-ready results, does not mean that an entire field is devoid of value or not science. There are other fields, such as pharmacology, which likewise have troubling corrupting influences, and yet can also be real science that produce real value.

Anyhow, we've diverged pretty far from the original topic, as the original topic is not even within the field of social psychology.

  you haven't the slightest idea what makes science science.

  you have no idea what you're talking about. Not a clue

  you do not know what it is. How do I know this? You 
  defended social psychology for over four hundred words 
  without even mentioning it.
Please refrain from the ad-hominem attacks. They bring the level of the discussion down considerably.

I do not mention social psychology because this research is not social psychology.

This research is founded on well known, rigorous research on the effects of lead poisoning.[1] This is not some surprising, out-of-left-field result which came out of the blue from some small study on a handful of college undergrads, which is the problem with many psychological and social psychology findings; this is based on building on many years of research and meta-analysis of epidemiological research. The original paper on this topic[2] was written in 1999, follow up papers from other authors have been published in 2007[3].

There have been further studies confirming this effect; this 2008 paper in PLOS Medicine[4] measured prenatal blood lead levels in pregnant mothers, childhood lead levels, and later correlated that with arrest rates. They too found a strong correlation. Other studies on the same cohort have shown a strong correlation between blood lead levels and reduced brain mass[5].

Here's a paper that compares several of the hypotheses that you have brought up[6]. Now, I take this paper with a large grain of salt, because it clearly comes from a biased source, the Heartland Institute, which is clearly biased towards libertarian findings; and the point of the paper is to argue against drug prohibition, or excessive enforcement of it (I am sympathetic to that viewpoint, but view their research on the topic with a heavy dose of skepticism). However, they compared several of the proposed determinants of violent crime rates, such as incarceration rates, access to legal abortion, environmental lead, and so on, and the only one of them (outside of the one that they're obviously promoting, drug prohibition) which held up with a consistent correlation was environmental lead.

[1]: https://en.wikipedia.org/wiki/Lead_poisoning

[2]: http://www.ricknevin.com/uploads/Nevin_2000_Env_Res_Author_M...

[3]: http://www.nber.org/papers/w13097.pdf?new_window=1

[4]: http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fj...

[5]: http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fj...

[6]: http://news.heartland.org/sites/default/files/c11845.pdf


> The exact same is true of this topic.

No, this is false. Try to imagine a conclusive falsification of the claim that's being made in the linked article. Falsification doesn't mean casting doubt on a hypothesis, it means proving it wrong. In fields such as the one under discussion, with p-values typically in the range 0.01 to 0.05, nothing is ever falsified. Some ideas are abandoned through embarrassment, but none of them are ever falsified.

If I say all swans are white, someone can falsify my claim by locating a black swan. If i say that a reduction in the crime rate might have a cause-effect relationship with a reduction in the lead content of gasoline, that's a perfect article for social psychology, because it's not possible to argue against it in a scientific sense.

> Please refrain from the ad-hominem attacks.

My posts aren't being downvoted because of their tone, they're being downvoted because my position is correct and it makes people uncomfortable to hear that so much of modern science ... isn't science.

Apropos, here's Richard Feynman making the same points, for the same reason, 40 years ago: http://neurotheory.columbia.edu/~ken/cargo_cult.html

> the only one of them (outside of the one that they're obviously promoting, drug prohibition) which held up with a consistent correlation was environmental lead. [emphasis added]

I shouldn't have to say this, but correlation, however consistent, doesn't equal causation, and causation -- testable, falsifiable theories -- is the foundation on which science is built.

Ask yourself why Creationism isn't taught in public school science classrooms. This is so because Creationism isn't science. That, in turn, was established in a trial -- several, actually -- in which the court's decision was based on the fact that science was shown to require falsifiability, not mere argument. Here's one such court ruling:

http://en.wikipedia.org/wiki/McLean_v._Arkansas

Quote: "The judgment defined the essential characteristics of science as being:

   * It is guided by natural law;
   * It has to be explanatory by reference to natural law;
   * It is testable against the empirical world;
   * Its conclusions are tentative, i.e. are not necessarily the final word; and
   * It is falsifiable."
Think about this. If your position were accurate, the Creationists would have the right to present Creationism -- argument without falsifiability -- as science in public schools.


  No, this is false. Try to imagine a conclusive 
  falsification of the claim that's being made in the 
  linked article. Falsification doesn't mean casting doubt 
  on a hypothesis, it means proving it wrong. In fields 
  such as the one under discussion, with p-values typically 
  in the range 0.01 to 0.05, nothing is ever falsified. 
  Some ideas are abandoned through embarrassment, but none 
  of them are ever falsified.
Are you trying to say that the only work that can ever be falsified is that which is black and white, cut and dried, you will just be able to pull a counter example out an someone will go home in shame because their pet theory was disproven?

That's not how the majority of science works. The majority of science is statistical, and noisy, with lots of confounding variables. That doesn't mean that it's not falsifiable; just that it's messy, it doesn't follow some neat perfect model of how science is supposed to work.

This topic, whether environmental lead leads to increased crime, absolutely is falsifiable, just as falsifiable as the laws of thermodynamics, which are likewise statistical in nature. Now, the laws of thermodynamics are a lot easier to test, and you can get much bigger sample sizes much more easily; but that doesn't have any bearing on whether a theory is falsifiable, just how easy the process is.

Lack of falsifiability means that a theory makes no predictions. It means that there is no way to distinguish a world in which the theory is true, from one in which it is not.

  If i say that a reduction in the crime rate might have a 
  cause-effect relationship with a reduction in the lead 
  content of gasoline, that's a perfect article for social 
  psychology, because it's not possible to argue against it 
  in a scientific sense.
Of course it's possible. You would just need to show sufficient evidences of cases of long term, widespread (population wide for a large enough population) increases in environmental lead where the crime rate did not increase 20 years later, or likewise widespread decreases in environmental lead where it did not decrease 20 years later. If you managed to come up with enough data of that sort, without having selectively chosen the data solely for that purpose (or with some other biased methodology that caused you to prefer data that prefers one result over another), that would be a falsification.

Other possible falsifications would be to test blood lead levels on a controlled population, and then observe that population's tendency towards criminal activity. If environmental lead were strongly correlated, but blood lead levels were not, then that might indicate that the causation was not ingested lead, but some other variable that was predicting both, such as industrial activity or development levels.

  My posts aren't being downvoted because of their tone, 
  they're being downvoted because my position is correct and 
  it makes people uncomfortable to hear that so much of 
  modern science ... isn't science.
There is a lot of modern science that is flawed. On the other hand, there is a lot of science from any era that is flawed. Pseudo-science has existed as long, if not longer, than science. No one is afraid to admit this. I am well aware of poor science, junk science, and pseudo science, as are many people in this thread. You have provided a lot of arguments against junk science in general, but have not provided one bit of evidence that this study is junk science.

I don't downvote, I prefer to make my points explicitly in writing (unless it's just someone who is a blatant troll). You have been behaving poorly in this thread; trying to claim that I don't know what science is, rather than actually addressing my points. You have been criticizing the research by pointing out completely different fields which produce dubious results for sensation headlines (epidemiology and social psychology are quite different). You seem to be arguing from emotion, as you have some kind of prejudice, possibly well deserved, against certain fields of study; but you're now applying that against other, completely unrelated research, instead of actually engaging with the science to try to learn something about how the world works.

  I shouldn't have to say this, but correlation, however 
  consistent, doesn't equal causation, and causation -- 
  testable, falsifiable theories -- is the foundation on 
  which science is built.
Falsifiability does not imply that you have established causation. A correlation is just as falsifiable as a causal relationship; all you have to do is show sufficient evidence of the same two variables without the given correlation. Likewise, you can have completely unfalsifiable causal statements; like "God made the heavens and the earth."

You are drifting further and further from actually engaging the topic at hand; all you are doing is rambling about some of your pet peeves about bad science and pseudo-science, but you haven't established that the study in question has anything to do with them.

  Think about this. If your position were accurate, the 
  Creationists would have the right to present Creationism 
  -- argument without falsifiability -- as science in public 
  schools.
I have no idea how you are getting a lack of falsifiability from this study at all, and thus no idea how you're linking it to creationism. Creationism doesn't make predictions. This study does. Those predictions can be tested. Yes, they are statistical in nature, and subject to confounding factors and measurement error, but that's the nature of science. Feynman's example of the oil drop experiment was likewise subject to similar kinds of confounding factors and measurement errors; luckily, that was one that was a lot easier to collect data on so it could be much more easily refined.

Would you say that the relationship between smoking and cancer is bad science? The science there has all of the same issues; there are ethical issues with asking someone to smoke in order to determine causation, so all you can do is population wide observational studies and observe a lagged rate of cancer among a population that is strongly correlated with the smoking rate 20 years earlier, and the higher rate of cancer among those who smoke than those who don't. The combination of lagged effect, strong correlation, and explanatory science based on known mechanisms of acute toxicity can all combine to make a fairly compelling argument.


That's pretty disappointing; a big formatting change like that means that it'll increase friction of pulling fixes and changes between the two versions.


But it makes it readable, which makes it easier to find bugs. This is vastly more important.


This, very much.

Normally I try to avoid getting into arguments about style and such, it's just bikeshedding. But if the OpenBSD people are going to clean up, review, audit, fix, improve and maintain this code, it makes all the sense to switch to a style they are used to and comfortable with. It also makes sense to nuke all the junk that gets in the way and which they will not and cannot test because it is not at all relevant to the platform.

Besides, the OpenSSL code is a fucking mess.


At some point you need to say, "This code is a mess, and pretty much anything we do to it will be an improvement."

I spent a couple hours with the OpenSSL source last week, and agree that a major cleanup is necessary. This means

- Rip out unused stuff - Rip out stuff that obviously doesn't work - Make big cleanups

I've done this in the past with token-level tools to ensure that I wasn't breaking things.

OpenSSL desperately needed to have a bunch of garbage removed. It needs good tests, it needs to be slimmed down and made clean so that the code is readable and maintainable.

Frankly I wouldn't trust fixes to the old version anyway, given its state.


> At some point you need to say, "This code is a mess, and pretty much anything we do to it will be an improvement."

In software development it's often the case that the first claim is true, but the second isn't.

> Frankly I wouldn't trust fixes to the old version anyway, given its state.

It will be interesting to see if this get organized and properly funded, given this library's importance to international commerce.

> Frankly I wouldn't trust fixes to the old version anyway, given its state.

Clean slate! An admirable goal, and clearly the best solution. Let's see if this actually happens.



Maybe because people frequently reference scrypt as a password hashing function, but when you download the source and look in the top level, what you get is an example of an encryption utility and no documentation other than a man page for said encryption utility?

Taking a look at the distribution, I can imagine being confused, as there's nothing to indicate that crypto_scrypt is what you want to be using.

Remember, not everyone who uses your library is a security expert, or will spend the time to read all of the code; they may be someone who has heard "you should use scrypt, it's the strongest readily available password hashing function", so they downloaded it and built it, but may not know where to go from there, so they take the most visible code, main.c, and use that as their example starting point.

Here's an example of something you could write in a top-level README to make this a little more clear. Feel free to use if you feel like, or reword as necessary. The main point is just to point people who are looking for a password hashing function to the right place, as that is what a lot of people looking at this may be looking for, and may not find immediately from the example given.

  This project provides the scrypt utility for performing password based symmetric 
  encryption, as well as libscrypt for using the functionality in your own software. 
  See scrypt.1 for details on use of the scrypt encryption utility.

  libscrypt implements the scrypt PBKDF, which can be used to derive a symmetric 
  encryption key for a password, or can be used as a password hashing function for 
  storing hashed passwords in a form that will make them more difficult to guess via 
  brute force.

  To encrypt and decrypt files or buffers using a symmetric cipher with a key derived
  using the scrypt PBKDF, use the functions declared and documented in
  lib/scryptenc/scryptenc.h.

  To implement a password hash, or use the scrypt PBKDF directly to derive a key for
  some other purpose such as use with a different symmetric cipher, use crypto_scrypt
  defined in lib/crypto/crypto_scrypt.h


Maybe because people frequently reference scrypt as a password hashing function, but when you download the source and look in the top level, what you get is an example of an encryption utility and no documentation other than a man page for said encryption utility?

You mean the code which is labeled on the website as "A simple password-based encryption utility is available as a demonstration of the scrypt key derivation function"?

Remember, not everyone who uses your library is a security expert, or will spend the time to read all of the code

I know people won't read all the code... but you'd think that people would at least read the comment directly above the function prototype

    * Encrypt inbuflen bytes from inbuf, writing the resulting inbuflen + 128
    * bytes to outbuf.
and realize that this doesn't even sound remotely like what they were looking for.

sigh

I've added a warning to the top of scryptenc.h which will hopefully help to point people in the right direction.


  You mean the code which is labeled on the website as "A 
  simple password-based encryption utility is available as 
  a demonstration of the scrypt key derivation function"?
Think of this from the perspective of someone who's looking for a password hashing function. Remember that a lot of people are fairly unclear on the idea of password hashing vs. encryption; lots of people say things like "the password database was encrypted" when they mean "the password database was hashed." For someone who is a bit fuzzy on the idea, but has been pointed to scrypt as the solution for secure password storage, they may just think "oh, it encrypts the passwords, OK" and go ahead and use the encryption functions.

The additional note is definitely helpful; I still think it would be good to put something at the top level which points people in the right direction.

As is, if someone wants to figure out what to do for password hashing, they have no documentation at the top level. The most promising thing to try, main.c, just uses the encryption functions. If they then jump to the definitions, your note is over 50 lines above the declaration of scryptenc_file, well outside of what will normally show up in their editor window. Yeah, they will probably have to scroll up a little to find the definitions of parameters and return values, but it's possible to do that and miss the note.

I know, people should read the whole documentation, but it's probably easier to just point them in the right direction from the start than to convince people to read the full documentation.


There are tons of bugs which are easy to describe in prose, and for which such an explanation can easily allow someone familiar with the codebase to be able to reproduce or even fix it right off the bat, but which would be quite difficult or even impossible to create an automated test for that could reproduce it in a way that the bug tracker could check.

It might be nice for bugtrackers to offer an optional system that allows you to submit an automated test case, which could be run in some kind of sandbox, and bugs that were easily reproducible would be marked as such and could get more attention. But anything that simply discarded all other bugs would be throwing the baby out with the bathwater; you would simply not hear from your users as much, on lots of important issues.


> Why was "The Outreach Program for Women" allowed to so significantly impact Gnome's primary mission?

One of the Gnome Foundations primary missions is "Fund Receipt and Disbursement": "The foundation will be in charge of disbursing these funds to the benefit of GNOME and, to the extent possible, in accordance with the wishes of the benefactor" (https://wiki.gnome.org/FoundationBoard/Resources/Charter).

The funding of the OPW sounds like it's completely appropriate here. Several benefactors and organizations have wanted to fund this activity, and the Gnome Foundation provides the means for them to do so.

> Why is the Gnome Foundation is funding and administering this program to begin with?

It is a multi-organizational program, between several different projects. It looks like rather than setting up a whole new organization just for this one program, they decided to have one group do the organization and running of the program, while the other groups contribute money to that one. It can be expensive and inefficient to set up too many new organizations; there is a lot of overhead to a new organization.

It looks like the problem is that because the OPW funds internships, the Gnome Foundation winds up having to pay them at a steady rate, while it may get payments from the other organizations in lump sums. It sounds like the Gnome Foundation has not been keeping up with invoicing the other organizations, so while they'll be in the black once everyone settles up, right now they've run short of cash on hand due to all of the outstanding balances.

> it absolutely should not financially impact Gnome's ability to ship Gnome.

Is shipping Gnome the job of the Gnome foundation? As far as I can tell, the foundation is there to organize conferences, run the server infrastructure, and run programs like the OPW. Actual development is mostly done by member companies, like Red Had and SUSE.

> Given that the outreach program was started in response to the lack of women candidates for Google's Summer of Code, this program could (and given fiduciary responsibilities, should) have focused solely on non-exclusionary outreach to women Summer of Code candidates.

That's not necessarily the best way to address the problem. It may be that certain aspects of how Google's Summer of Code works inherently biased towards men; or at least, people who are already fairly comfortable in the community and with open source software, which given the current substantial gender imbalance in the community, can mean that it's not really a suitable way to encourage more women to get involved.

Running the internships separately from the Google Summer of Code is perfectly fine; they just need to do a better job of managing the finances.


> One of the Gnome Foundations primary missions is "Fund Receipt and Disbursement": "The foundation will be in charge of disbursing these funds to the benefit of GNOME and, to the extent possible, in accordance with the wishes of the benefactor"

This provides its own rebuttal. Running out of funds necessary to pay for core commitments is not "to the benefit of GNOME", and certainly not in-line with their charter.

To quote the same charter -- specifically the mission statement:

    The GNOME Foundation will work to further the goal of the
    GNOME project: ___to create a computing platform for use by the general
    public that is completely free software.___ (emphasis mine)

    To achieve this goal, the Foundation will coordinate releases of GNOME
    and determine which projects are part of GNOME. The Foundation will act
    as an official voice for the GNOME project, providing a means of
    communication with the press and with commercial and noncommercial
    organizations interested in GNOME software. The foundation may produce
    educational materials and documentation to help the public learn about
    GNOME software. In addition, it may sponsor GNOME-related technical
    conferences, and represent GNOME at relevant conferences sponsored by
    others, help create technical standards for the project and promote
    the use and development of GNOME software.
Gender-exclusionary social programs only very loosely fall under this mission statement, and only given an exceedingly broad interpretation of the mission.

[edit]

I think this admission provides sufficient evidence of charter overreach:

    The changes were not tracked effectively from the
    point when other organizations joined the OPW. This
    impacted not only our ability to manage the OPW 
    administration, but also to keep up with the core 
    financial tasks of the Foundation -- tasks which
    already needed the full attention of the 
    Foundation's employees and the board.
This appears to be no small matter of malfeasance by virtue of clear overreach beyond their charter. I can't help but wonder if there has been some form of social cause "regulatory capture" within the foundation that drove this surprising decision.

A recall referendum, as defined by the charter, seems warranted.


  This provides its own rebuttal. Running out of funds 
  necessary to pay for core commitments is not "to the 
  benefit of GNOME", and certainly not in-line with their 
  charter.
Of course running out of funds is not to the benefit of GNOME. The running out of funds part was an accident; no one wanted to do that. If there were any other program that had similar characteristics; something where the foundation were paying out a lot of money with the expectation of being reimbursed, but were not reimbursed as fast as they had expected, it would have had the same effect.

I think you need to separate your distaste for the program in the first place, from the immediate problem that they ran into a cashflow problem running it.

  to create a computing platform for use by the general 
  public that is completely free software...
  promote the use and development of GNOME software.
An internship program tailored towards a demographic that is traditionally heavily under-represented in both the users and developers of the free software eco-system sounds like exactly what the Gnome Foundation should be doing. Note the "general public" part; in order to reach the general public, and not just a fairly small niche market of existing free software enthusiasts, they need to broaden their appeal to demographics which don't traditionally use or work on the project.

Remember, this money is being used directly for the development of Gnome; these internships are for development of Gnome and other components of the free software stack, like the Linux Kernel, Firefox, and so on.

  This appears to be no small matter of malfeasance by 
  virtue of clear overreach beyond their charter.
Whoa, whoa. That's a rather strong accusation; you are accusing them of willfully harming the foundation against its charter?

The foundation has been running this program for several years, and the membership has never objected to it. It's pretty clear that the membership does not believe this program to be at odds with the mission of the foundation.

Now, obviously the financial mismanagement part is a problem, which the board of directors should be working on fixing. But that doesn't mean that you need to throw the whole program out, just be more careful about working on cashflow.

  Is there a mechanism by which the board can be recalled? 
  It does not seem at all unwarranted.
Are you a member of the Gnome Foundation? If not, why do you care so much about how it operates, and sets its priorities; or if you care so much, why don't you become a member and get involved?


Doesn't teacup50 have a fair point? I would say that at least Mediawiki, Openstack, and some robotics foundation (all participating in OPW by taking interns) are well out of scope. All GNOME gains from being in charge of it all is a chance to make mistakes.


> Whoa, whoa. That's a rather strong accusation; you are accusing them of willfully harming the foundation against its charter?

If I understand what happened here correctly, yes. The OPW program was questionable within the charter, but this went beyond administering OPW for Gnome itself.

The board seemingly elected to administer the program for external organizations, to the direct detriment of their ability to manage the Gnome Foundation's responsibilities.

This outcome was obviously unintentional, but the choices that led to it were not.


Wow, a wall of text. But all I can read is a seriously problematic and classic case of mansplaining.

Seriously, Hacker News can be so misogynistic at times it's not even funny.


As a reader, I far prefer this to any kind of summary. This was well written, thoughtful, and thorough. I can't imagine any value that any summary would have added to this, unless I were just hoping to skim it over my lunch break. And I greatly prefer to read one well written, thoughtful, thorough article in a day, then a dozen summaries which inflame the senses without providing any particularly useful information.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: