Slashdot Mirror

That's about the size of it. Damned if you do, damned if you don't. I believe they are working on a new type that is pollen sterile so that no cross pollination take place to begin with (could be wrong here though, I don't know much about that particular trait) and what are called 'traitor seeds' where the seeds will not grow without an application of a certain chemical. This is one of those things that really highlight how it is not about the properties of the GMOs themselves, but the fact that they are GMO. The issue that many of the anti-GMO crowd have with GMOs, once you rebut the nonsense and get down to the nitty gritty, is simply that they were made with biotechnology, not for any really rational reason. It's basically just an appeal to nature.

Also, needing to buy new seeds every year is nothing new. Farmers have been doing that since hybrids from seed companies came on the seen in the early 1900s, so the claim that GMOs mean you can't safe seed is not very relevant (some people, however, even encourage people to save their GMO seeds). There is something to be said for saving seed if you have a backyard heirloom garden like I do, but if you want to get the most out of your plants, that increasingly means some sort of GMO, and that is hopefully going to mean some sort of safeguard against genes escaping into the wild. I mean, I haven't heard a whole lot about how much that actually happens in practice, but still, it is something to keep in mind, because the ecology of GMOs is easily the most scientifically complex and controversial aspect of them because you must weight the danger posed to escaped genes with the benefits they provide and damage they can prevent.

Please, try to learn something more about science before posting as if you understand it. When you say something like:

Mars has an atmospheric CO2 content that by volume and mass is greater than Earths. Why is Mars not hot? Why does the greenhouse effect not slip out of control there?

...you make yourself sound like a nitwit to those people (a majority of Slashdot readers) who know that Mars is further than Earth from the Sun. Similarly, when you say that:

The odds of IR radiation striking a CO2 molecule on the way up on Earth is extremely small. If this weren't true, IR pictures of fields and cities would be blurred by the scattering caused by CO2.

...you sound outright disingenuous. The objects in an IR picture are usually within a hundred metres or so. The atmosphere between the Earth's surface and space is equivalent to about 7 km of air at sea-level density. If you took an IR picture from this distance, it would be blurred. Actually, this is the reason why IR astronomy is done with satellites instead of ground-based telescopes - IR from an astronomical object doesn't reach ground level without being scattered. (Refer to, eg, this graph. The vertical axis is the fraction of radiation that is blocked by the atmosphere. Notice that it goes above 50% for several bands within the IR range.)

You are correct to say that the sort of name-calling that debates on this subject frequently degenerate to is not helpful, but neither is loudly expounding a viewpoint based on an incomplete understanding of the facts. (There's nothing wrong with being uninformed - but when you pretend to be informed, you lead to the misinforming of others.)

My biggest concern with climate modeling right now is that climate scientists are not the equal of computer scientists, and this gives one pause [wikipedia]

My problem is with people who take Wikipedia as an authority, and not as a starting point.

Even back in 2004, the models were using adaptive grids (e.g., http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.60.5091 ) where it makes sense to do so. And I don't think anybody uses constant latitude-longitude grids any more-- this one http://www.ncl.ucar.edu/Document/Graphics/contour_grids.shtml shows some of the more common grids.

It's often possible to determine how a pseudorandom numbers were generated by finding characteristic statistical biases. See http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.19.7206. The authors found significant biases under some tests for true randomness in a number of standard encryption algorithms, meaning that one could run these tests and likely distinguish between an AES encrypted bitstream and one generated some other way.

Perhaps this classic paper might be worth a read to him. The figures are hilarious: "Finding Naked People", by Margaret Fleck, David Forsyth, and Christoph Bregler http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.47.8929

They do exist, and Macondo isn't unique as an oil field either. There is ample evidence for natural plumes of oil and gas seeping from the sea floor at many locations world-wide, including the Gulf of Mexico and offshore California where they have been well-studied. There are some details at this site, which includes a number of publications about California seeps. This site has more information, including a nice aerial photo of the natural oil slick produced by the seeps off the coast of California. This paper [PDF] is a good summary of the best known California locations.

here and here [PDF] are some examples [PDF] of seeps, gas plumes [PDF], and seep-related life [PDF] in the Gulf of Mexico [PDF]. The tube worms growing on asphalt [PDF] or "ice worms" in burrows in gas hydrate (!) [PDF] are particularly cool. Some life *likes* oil and gas leaking into the ocean.

The bottom line is, seeps of gas and liquid hydrocarbons into the water column happen all the time, and the ocean deals with them by bacteria eating the oil. On the sea floor there are flourishing biological communities associated with the release of hydrocarbons, like a little "oasis" of life in the deep sea, supported by creatures eating the bacteria that are in turn eating the hydrocarbons that are expelled. However, the rate of release at the BP well is several times the total output of natural seeps across the entire Gulf of Mexico, so the scale of the release is much bigger and concentrated in one gigantic point source. It would be like trying to feed on a volcano. There certainly won't be any seep communities setting up at the BP well any time soon :-) Anyway, the life in the ocean will consume this stuff as it spreads out, whether it is in a plume or on the surface, but it will take a while. Also, the plume they are talking about in the water column is extremely low concentration (ppb average). 99%+ of the oil is making it to the surface, and most of the gas is either dissolving in or venting to the atmosphere. The plume is interesting from a scientific perspective and probably will have some kind of environmental effect that could be important, but it's not the main part of the environmental problem. People are hyped about it because they are obsessed with the idea that the disaster could be 10x worse than the oil that is obvious on the surface. These studies show that simply isn't the case because of the low concentrations. A significant fraction of the oil is not lurking below, only a tiny, instrument-detectable amount is.

No, it means that in one of the few examples of a laissez-faire market in the modern world, Veblen was right. No matter what the economic system, the main engine of expanding commerce, inventors, get fucked.

(For those interested in original text, I would note that all his major works were published in the late period of the public domain, including The Theory of the Leisure class (pdf).

relatively low power lasers can apply enough force on a (dead) satellite or other orbital trash to gradually reduce their orbit until they eventually burn up:

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.120.6304&rep=rep1&type=pdf (DARPA/USAF paper from 2000)

the trick is not blinding other satellites in the process.. especially 'secret' ones that you didn't know were there... also, if the satellite is not really 'dead' (just nonresponsive), it may continue to use it's own OMS etc to counter the laser pressure and stay in orbit... at least until it runs out of fuel.

I was aware of the using-diffraction-to-compute-Fourier-transforms idea; in fact, I was under the impression that it was somewhat popular before the advent of digital computers. A really good comparison.

Still, I think that maybe "cheating" is exactly what we should be doing more of. We can use obscenely-sophisticated multigrid PDE solvers to solve Navier Stokes... or we can build a wind tunnel and instrument it with sensors. What I'm wondering is whether there are other physical processes that are good analogues for different important problems.

One which is particularly important is the Hamilton-Jacobi-Bellman equation -- the PDE whose solution is essentially the holy grail in Optimal Control. If we had hardware to solve that quickly, it'd really do wonderful things for robotics and AI. One can even imagine solving it in 2d by varying the index of refraction in a material. However, in 2d it's not that hard to solve on digital computers either. The problem is that the complexity of solving it grows exponentially with the dimensionality of the state space (Bellman's "curse of dimensionality"), and I think it's very non-obvious how one might go about building an analog computer to solve it in dimensions higher than three.

Another (which is often a "good enough" approximation to HJB) is Laplace's equation. Physical analogues for that are extremely common... RC networks, thermal systems, global illumination... so this could be a good way to e.g. generate robot navigation functions (see e.g. [1]). IIRC there are even people building analog circuits to do exactly this; I find that rather cool (anybody know who it is that's doing this?).

A third example -- this one an ODE rather than a PDE -- which is quite cool (though it doesn't seem super useful) is [2], which can among other things sort lists (for this purpose a bitonic comparator network seems more practical though). Still, very cool.

Of course, you also get all the problems that go along with analog computation: component drift, noise, etc. I wonder if these can be alleviated by (1) controlling the environment (e.g., temperature control), and (2) using some slower digital systems in adaptive control loops to counteract drift. I'm sure that the analog electronics guys have considered both of these ideas, and there are probably papers on them! (I'm aware of e.g. transistor matching, which is standard practice...)

I guess my basic point is just that I think it might be fruitful to continue looking to the physical world for systems that naturally do the computations we care about. It might not be as general-purpose as a Turing Machine, but if it's a problem that matters enough it can become a coprocessor.

[1] C.I Connolly, J. B. Burns, and R. Weiss. Path Planning Using Laplace's Equation. ICRA, 1990. (PDF.)

[2] R. W. Brockett. Dynamical systems that sort lists, diagonalize matrices, and solve linear programming problems. Linear Algebra and its Applications, 1991. (PDF.)

The inhospitality of England to vineyards has been greatly exaggerated. Nevertheless, the existence of a regional warm period in northern Europe is not really in dispute (PDF)

Furthermore, it seems to suggest that the Earth can warm up several degrees and actually be beneficial for mankind in terms of increased growing seasons for many areas and increased food production in general. It sort of begs the question.... what are we worried about even if the global environment is warming up?

It is certainly possible that some northerly areas might benefit from longer growing seasons. On the other hand, the temperate areas that currently enjoy a near optimal climate, and that produce much of the world's food would be expected to suffer. So a country like Greenland might benefit, while the US would probably lose big. Since agriculture cannot shift from one region to another that rapidly, countries that currentl import large quantities of foods like grain may experience massive starvation.

A great part of the problem is that head smacking obvious issues like what caused the medieval global warming

The existence of a medieval global warming period seems to be an article of faith among opponents of global warming, even though the evidence that it was global rather than regional is much weaker (PDF) than the evidence for modern global warming. The oddest thing is that they seem think that the possible existence of some additional mechanism that is not understood whereby the earth could warm more than is expected from current global climate models models should make people less concerned about the possible consequences of modern global warming. If the medieval warm period was indeed global, it would argue that there is some additional mechanism that could add to or amplify the modern CO2 induced warming so as to cause global temperatures to shoot up even higher than projected.

Here's the data behind one of his most recent papers. Note that he's included his Matlab code.

Is what I meant to say. Clearly there's a conspiracy to keep this data from your oh so capable hands!

(I hope you realise that most of your post consists of the well known strawman fallacy)

Oh good, so if I was setting up strawmen, then you aren't claiming that Mann et al 2008 used tree ring data improperly? Because that's the only position I attributed to you, as far as I can see. It would explain why you never supported that position; you never held it in the first place, you just insinuated that it was true.

Baillie, and many other dendrologists, are very wary about using tree rings as temperature proxies. I.e, your claim that Baillie's comment is about Keenan specifically is not correct, it's about oaks - no matter who uses them. (Climate signal in tree-ring chronologies in a temperate climate: A multi-species approach. Suarez, Butler, Baillie, 2009)

Did you even read the paper you're citing here? Did you even read the abstract? For those of you with access to it, here's a link to a pdf of the actual paper (it might be public access, I don't know).

Here's a quote from the abstract:

We nd that the moisture-related parameters, rainfall and the Palmer Drought Severity Index (PDSI), and to a lesser
extent, maximum and mean temperatures, can be reconstructed.

Then look on page ten. There's a heading. It's "Reconstructions". In that part of the paper, Suarez et al (and in this case, et al includes Baillie) proceed to reconstruct temperature data using tree ring data. So apparently, Baillie's statement is about oaks, no matter who uses them - including the Baillie himself. Indeed, the point of this paper is that by using multiple tree species you can extract okay data from just the dendrochronological record. Keep in mind that Mann et al weren't doing that - they were using the dendro record plus other proxies. This paper is about how far you can get with just dendro and some calibration.

As a side note, I would be wary of citing the Wegman report; the social networking segment, at least, is full of outright plagiarism which is kind of concerning.

Anyway, now that I look at Mann et al 2008(warning, pdf and also you might not have access to it), it doesn't matter if the bristlecone data isn't that great. Here's a quote from the paper:

For both methods, we perform reconstructions both with and without dendroclimatic proxies to address any potential sensitivity of our conclusions to issues that have been raised with regard to the reliability of tree-ring data on multicentury timescales(4,11,16,19,33,34).

And guess what? It doesn't really make much of a difference. Here's (pdf again) a graph of the various reconstructions both with and without tree ring data; the no tree lines spike a bit more, but are generally consistent.

Funny how the website you linked to never even mentions this fact. Weird, huh? It's almost like they (and you) didn't read the actual paper.

Oh yeah and if you're wondering how they know that Mann et al has exactly 119 oak tree records in their reconstruction, it's because the stuff is available online. There's a link in the paper.

• Penn State
• the British Parliamentary inquiry
• the Oxford panel
• Nature.com

Accused and exonerated.

...

WRONG!!!!!

Straight from the horse's mouth:

Inquiry into climate scientist moves to next phase

...

In looking at four possible allegations of research misconduct, the committee determined that further investigation is warranted for one of those allegations.

...

In the investigatory phase, as in the inquiry phase, the committee will not address the science of global climate change, a matter more appropriately left to the profession. The committee is charged with looking at the ethical behavior of the scientist and determining whether he violated professional standards in the course of his work.

They ignored the science, and they ignored 3 out of 4 allegations.

This has been settled since FEBRUARY. http://www.research.psu.edu/orp/Findings_Mann_Inquiry.pdf Go to page 5.

http://www.writing.engr.psu.edu/teaching_slide_design.html http://www.writing.engr.psu.edu/slides.html btw, today i had to give a ppt to a fortune 500 company interested in us (small company smellign $$$$) i had one killer handdrawn figure, you could see the lightbulbs go on when people saw it

http://www.writing.engr.psu.edu/teaching_slide_design.html http://www.writing.engr.psu.edu/slides.html btw, today i had to give a ppt to a fortune 500 company interested in us (small company smellign $$$$) i had one killer handdrawn figure, you could see the lightbulbs go on when people saw it

Global climate reconstructions do not support the existence of a global medieval warm period. See, for example,
http://www.meteo.psu.edu/~mann/shared/articles/MannetalScience09.pdf

This is not what you asked for, but it is an article summarizing the results of a number of studies (with references) plus the article writer's own personal experience.

Here is a different study that looked at the differences between complex interruptions and simple interruptions during the execution of a complex task. Bear in mind that the "complex task" was nowhere near as complex as various programming tasks can be. They found a complex task interrupted by a simple task generally cost about 4 minutes to get back into the task, and a complex task interrupted by another complex task took close to 8 minutes to get back into the task. An interesting affect they noted, however, was that when a complex task was interrupted by another complex task, when the person went back to the main task they made fewer errors for a time. That was not the case for a simple interruption.

Isn't this why countries such as France have completely abolished juries, and judgments are rendered solely by Judges...Of course, that has its own problems, like the fact that one crappy judge will make a lot of bad verdicts

Generally, defendants can waive their right to a jury trial and have the case go before a judge, they think that is to their benefit.

It seems that government would certainly like to abolish trial by jury, as it has this annoying tendency to slightly slow the growth of the prison-industrial complex. In blatant contradiction to the Constitutional requirement, SCOTUS has somehow become illiterate regarding the phrase "all criminal prosecutions" and ruled that you don't have a right to a jury trial if the sentence is less than six months -- even if you're facing multiple counts and could spend years in jail. And the state continually tries to keep juries ignorant of their right to judge questions of law as well as of fact.

That is why most of the comp sci courses I've taken here at PSU have assignments like: "create a program that is at least 1000 lines and uses x, y, and z" or "Create a program that has a GUI using a text box, a button, and radio buttons and outputs data inside the GUI in some form" (well, more complete descriptions than those, but you get the idea). Much harder to cheat that way.

It seems they've been changing the courses since I've taken them - even in the past semester - but this is generally what the majority of my projects were:
http://php.scripts.psu.edu/djh300/cmpsc221/proj2-open.htm

Sure, there are homeworks too in some classes that are much smaller and generally easier to cheat on, but those are usually less than a quarter of the final grade. The rest is projects and exams.

PSPACE-Complete: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.52.41

Most languages still handle concurrency very badly. C and C++ are clueless about concurrency. Java and C# know a little about it. Erlang and Go take it more seriously, but are intended for server-side processing. So GUI programmers don't get much help from the language.

In particular, in C and C++, there's locking, but there's no way within the language to even talk about which locks protect which data. Thus, concurrency can't be analyzed automatically. This has become a huge mess in C/C++, as more attributes ("mutable", "volatile", per-thread storage, etc.) have been bolted on to give some hints to the compiler. There's still race condition trouble between compilers and CPUs with long look-ahead and programs with heavy concurrency.

We need better hard-compiled languages that don't punt on concurrency issues. C++ could potentially have been fixed, but the C++ committee is in denial about the problem; they're still in template la-la land, adding features few need and fewer will use correctly, rather than trying to do something about reliability issues. C# is only slightly better; Microsoft Research did some work on "Polyphonic C#", but nobody seems to use that. Yes, there are lots of obscure academic languages that address concurrency. Few are used in the real world.

Ada 2005's task model is a real world, production quality approach to include concurrency in a hard-compiled language. Ada isn't exactly known for its GUI libraries (there is GtkAda), but it could be used as a foundation for an improved concurrent GUI paradigm.

This book covers the subject quite well.

A big problem is the event-driven model of most user interfaces. Almost anything that needs to be done is placed on a serial event queue, which is then processed one event at a time. This prevents race conditions within the GUI, but at a high cost. Both the Mac and Windows started that way, and to a considerable extent, they still work that way. So any event which takes more time than expected stalls the whole event queue. There are attempts to fix this by having "background" processing for events known to be slow, but you have to know which ones are going to be slow in advance. Intermittently slow operations, like an DNS lookup or something which infrequently requires disk I/O, tend to be bottlenecks.

Most languages still handle concurrency very badly. C and C++ are clueless about concurrency. Java and C# know a little about it. Erlang and Go take it more seriously, but are intended for server-side processing. So GUI programmers don't get much help from the language.

In particular, in C and C++, there's locking, but there's no way within the language to even talk about which locks protect which data. Thus, concurrency can't be analyzed automatically. This has become a huge mess in C/C++, as more attributes ("mutable", "volatile", per-thread storage, etc.) have been bolted on to give some hints to the compiler. There's still race condition trouble between compilers and CPUs with long look-ahead and programs with heavy concurrency.

We need better hard-compiled languages that don't punt on concurrency issues. C++ could potentially have been fixed, but the C++ committee is in denial about the problem; they're still in template la-la land, adding features few need and fewer will use correctly, rather than trying to do something about reliability issues. C# is only slightly better; Microsoft Research did some work on "Polyphonic C#", but nobody seems to use that. Yes, there are lots of obscure academic languages that address concurrency. Few are used in the real world.

Game programmers have more of a clue in this area. They're used to designing software that has to keep the GUI not only updated but visually consistent, even if there are delays in getting data from some external source. Game developers think a lot about systems which look consistent at all times, and come gracefully into synchronization with outside data sources as the data catches up. Modern MMORPGs do far better at handling lag than browsers do. Game developers, though, assume they own most of the available compute resources; they're not trying to minimize CPU consumption so that other work can run. (Nor do they worry too much about not running down the battery, the other big constraint today.)

Incidentally, modern tools for hardware design know far more about timing and concurrency than anything in the programming world. It's quite possible to deal with concurrency effectively. But you pay $100,000 per year per seat for the software tools used in modern CPU design.

In the original netflix competition the data they did not release birthdays, zip codes or gender. Every movie and user was given a unique (presumably random) id. Essentially the data you had to work with was a bunch of tuples:

(movie id, user id, rating (1-5), date)

Netflix claims they even added some random noise (changing the dates/ratings a little bit) to preserve anonymity. Turns out even this isn't enough to guarantee anonymity...you can cross reference this data with IMDB to look for similar date/ratings patterns and re-identify a lot of the people.

See the paper: How to break the Anonymity of the Netflix Prize dataset

You may not like his politics or his prognostications, but Gelernter has made solid contributions to computer science, especially in the field of distributed information spaces. http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.113.9679

Cute!

However, people fuck up all the time. So when you came out of your mom, were you told "Don't worry, nothing will go wrong." when you started screaming :)?

On a more serious note, there are workplaces where you have to hit buttons to prevent accidents like having your hands chopped off. Any more refined and also intrusive means to prevent more complex fuck-ups will enable higher risk technologies. This could be installed on a voluntary basis (people go to war voluntarily) and a government under competitive pressure might just permit this and then some.

You see, we won't go from free will today to scull cap tomorrow. Today we may have some sort of automated observation techniques:

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.99.8216&rep=rep1&type=pdf

In between there will be more incremental improvements. Nobody will worry too much from step to step apart from some notorious naysayers, you know how these things work. I mean people accept torture nowadays, the scull cap has way more benefits than torture if used on a large number of people.

I don't really see how your link supports your assertions. It says: " Archimedes' work was translated into Arabic by Thbit ibn Qurra (836-901 AD), and Latin by Gerard of Cremona (c. 1114-1187 AD). " That does not say that Archimedes' work comes to us through Arabic, it just says that it was at one point translated into Arabic. Indeed, this source seems to contradict your position: a collection of Archimedes works in 9th-century Constantinople form the "basis of the texts we have today." Why would Greek scholars in the 9th century who didn't know much Arabic, if any, make a collection of Archimedes in Arabic, instead of Greek, when the Greek manuscripts were clearly still available in the 9th century (how else would bin Qurra have translated Archimedes into Arabic?)? I think I'd believe that that collection (which "is the basis of the texts we have today") which they are is mentioned here was in Greek, unless you can provide some evidence that it was not...

This page seem to contradict your statement about Euclid's elements. If a 4th century edition formed the basis of texts until the 19th century, then the texts until the 19th century were not based on Arabic, because there is no way that 4th century text was in Arabic. And then that text was improved with another Greek version, to form the Heiberg edition, which, so it says (in 1971), "still stands". There's no mention of Arabic.

Other responders have similar problems (Ptolemy's Almagest: http://www.ibiblio.org/expo/vatican.exhibit/exhibit/d-mathematics/Greek_astro.html). But one of the other responders mentions Hero of Alexandira, and appears to be partly right: of his 14 works listed in this article [PDF], 1 comes to us from Arabic (the Metrica), and 1 comes from Arabic with some Greek fragments (the Mechanica). Both seem to be significant works.

So I still think there is this myth that "lots" of Greek works are transmitted to us through Arabic, although I'm happy now to have some examples of significant texts (2 of Hero's 14) that do actually fit this model.

You talking about the 'climatgate' incident with Michael Mann?
Pennsylvania State Uni did an investigation into this, as they don't really want to deal with scientists that cheat and they found no errors in the science, so there.

Does your speaker have anything concrete with citations to back his assertions up, or is he happily dismissing one of the few genuine advances in software engineering in the last decade?

we found that the code developed using a test-driven development practice showed, during functional verification and regression tests, approximately 40% fewer defects than a baseline prior product developed in a more traditional fashion. The productivity of the team was not impacted by the additional focus on producing automated test cases. This test suite will aid in future enhancements and maintenance of this code.

-- http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.129.7992&rep=rep1&type=pdf

A Spring 2003 experiment examines the claims that test-driven development or test-first programming improves software quality and programmer confidence. The results indicate support for these claims

-- http://portal.acm.org/citation.cfm?id=949421

Experimental results, subject to external validity concerns, tend to indicate that TDD programmers produce higher quality code because they passed 18% more functional black-box test cases.

-- http://qwer.org/tdd-study

We observed a significant increase in quality of the code (greater than two times) for projects developed using TDD compared to similar projects developed in the same organization in a non-TDD fashion.

-- http://portal.acm.org/citation.cfm?id=1159733.1159787

My apologies for the rough and ready citations, I only picked the ones I could find on the first fucking page of Google search results.

Mann was cleared of most of accusations: http://www.research.psu.edu/orp/Findings_Mann_Inquiry.pdf

Two things to recall here. First, Pennsylvania State University has a strong interest in clearing Mann's name. The investigation was not unbiased. Second, Mann was not a member of the CRU nor responsible for the attempts to evade FOIA requests there. If PSU had bothered, they'd have found me clear of most accusations too for similar reasons.

Certainly. However, models were available to other researchers and most of data is (and was) free.

Not the code though. And the identity of the data sources for IP protected data?

Which data exactly do you need? I personally worked with CSIRO HADISST dataset for ice coverage. There is GHCN-Monthly for temperatures. And GPCC from Germans which are also the part of the conspiracy.

Let me restate my concern here since I wasn't clear the first time. There's as far as I know only three paleoclimate aggregations that estimate global temperatures before modern times. Are there more of them?

Also, CISL doesn't appear to aggregate data, it just provides it.

And anything (and I really mean _anything_) Steve McIntyre says is probably false and/or misleading.

So what? It's still criticism (and he has been other than "probably false and/or misleading" at times in the past, remember the "hockey stick" complaint?). You were claiming that the data hadn't been called into question even though here is an example where it has been.

"And as I noted, the most common way is to not be a party required by law to comply with the FOIA. Moving on, just because something is legal to do, which incidentally doesn't appear to be the case with the CRU FOIA requests"

Mann was cleared of most of accusations: http://www.research.psu.edu/orp/Findings_Mann_Inquiry.pdf

"doesn't mean you should do it. Hiding important data and computer models is unscientific as noted by the original poster."

Certainly. However, models were available to other researchers and most of data is (and was) free.

"Ok, I looked. Not sure what your point was supposed to be. As I understand it, there are three prime aggregators of paleoclimate data, the CRU, a NASA unit headed by James Hansen, and something similar in the NOAA (the last being the only one that isn't running some obvious unscientific agenda). If there are other aggregators, then maybe you could just mention them by name rather than throwing links at me?"

Which data exactly do you need? I personally worked with CSIRO HADISST dataset for ice coverage. There is GHCN-Monthly for temperatures. And GPCC from Germans which are also the part of the conspiracy.

A colleague here tells me that CISL also provides aggregated datasets: http://www.cisl.ucar.edu/dss/

And anything (and I really mean _anything_) Steve McIntyre says is probably false and/or misleading.

the non-technical guy who kills ideas and can't be reasoned with

aka "The Cage Match Negotiator". The term comes from the professional wrestling format where multiple wrestlers enter the cage, but only one exits victorious. The cage match negotiator will "win the argument" at any price and damn the torpedos. This type of person can be particularly destructive in a large organization where high level managers have almost unlimited power to alter or cancel projects at will.

Original press release by PSU

One of the researchers, Professor Mohsen Kavehrad, has also worked on 100Gbps transmissions over 100m long Cat7, for example, so these people are not noobs playing with laser diodes from DVD-drives. The article mentions that "the researchers chose to take a different approach using multi-element transmitters and multi-branch optical receivers in a quasi-diffuse configuration." This article gives slightly more details: "The challenges regarding attenuation and multipath distortion can be overcome by using multi-spot diffuse configuration and fly-eye reception ( Kavehrad & Yun, 1992 ). In this configuration, the transmitted beam is split in a control manner into several narrower beams by means of holographic beam-splitters. The narrower beams illuminate selected spots on a reflecting surface. Thus, path loss due to diffusion is reduced, and fly-eye receivers can use diversity combining techniques to increase signal-to-noise ratio."

What this means is that each receiver only works with one narrow beam which is unlikely to be affected by different path lengths in the same signal and the signal from multiple receiver pairs is aligned and combined to increase the signal to noise ratio.

To the guy who can't help but laugh at the notion of pulsing a little light on and off at 1GHz: That's exactly what they're doing. The magic is in the receiver.

Original press release by PSU

One of the researchers, Professor Mohsen Kavehrad, has also worked on 100Gbps transmissions over 100m long Cat7, for example, so these people are not noobs playing with laser diodes from DVD-drives. The article mentions that "the researchers chose to take a different approach using multi-element transmitters and multi-branch optical receivers in a quasi-diffuse configuration." This article gives slightly more details: "The challenges regarding attenuation and multipath distortion can be overcome by using multi-spot diffuse configuration and fly-eye reception ( Kavehrad & Yun, 1992 ). In this configuration, the transmitted beam is split in a control manner into several narrower beams by means of holographic beam-splitters. The narrower beams illuminate selected spots on a reflecting surface. Thus, path loss due to diffusion is reduced, and fly-eye receivers can use diversity combining techniques to increase signal-to-noise ratio."

What this means is that each receiver only works with one narrow beam which is unlikely to be affected by different path lengths in the same signal and the signal from multiple receiver pairs is aligned and combined to increase the signal to noise ratio.

To the guy who can't help but laugh at the notion of pulsing a little light on and off at 1GHz: That's exactly what they're doing. The magic is in the receiver.

Original press release by PSU

One of the researchers, Professor Mohsen Kavehrad, has also worked on 100Gbps transmissions over 100m long Cat7, for example, so these people are not noobs playing with laser diodes from DVD-drives. The article mentions that "the researchers chose to take a different approach using multi-element transmitters and multi-branch optical receivers in a quasi-diffuse configuration." This article gives slightly more details: "The challenges regarding attenuation and multipath distortion can be overcome by using multi-spot diffuse configuration and fly-eye reception ( Kavehrad & Yun, 1992 ). In this configuration, the transmitted beam is split in a control manner into several narrower beams by means of holographic beam-splitters. The narrower beams illuminate selected spots on a reflecting surface. Thus, path loss due to diffusion is reduced, and fly-eye receivers can use diversity combining techniques to increase signal-to-noise ratio."

What this means is that each receiver only works with one narrow beam which is unlikely to be affected by different path lengths in the same signal and the signal from multiple receiver pairs is aligned and combined to increase the signal to noise ratio.

To the guy who can't help but laugh at the notion of pulsing a little light on and off at 1GHz: That's exactly what they're doing. The magic is in the receiver.

Yea, I'm at UP. The literature says 4GB, but it's been increased, they just haven't updated that yet.

http://www.collegian.psu.edu/archive/2009/12/27/university_doubles_bandwidth_l.aspx

Also go to the rescom 'bandwidth used' page - the scale now goes up to 10, rather than 4.

Expert or no expert, let's stick to the facts. Here's what wikipedia's page on fast deterministic tests for primality says:

The elliptic curve primality test, which actually proves that the given number is prime, can be proven to run in O((log n)6), but only if some still unproven (but widely assumed to be true) statements of analytic number theory are used. It is one of the most often used deterministic tests in practice.

The elliptic curve primality test is 1) 'a general-purpose algorithm, meaning it does not depend on the number being a special form' 2) has been able to produce a certificate for 'the 20,562-digit Mills' prime' in about 6 years of single CPU run time 3) as of 1991 'it is now possible to test arbitrary integers up to 400 digits in a few days on a single SUN 3/60 workstation' according to the Atkin and Morain paper Elliptic Curves and Primality Proving.

Lastly, from the AKS primality test page, it would appear that the AKS test is more significant as a theoretical result than as a practical algorithm.

I am not an expert in this field, however it appears to me that you may not be one either.

Possible specific impulse for an antimatter based star ship?

http://www.engr.psu.edu/antimatter/introduction.html

10^3 - 10^6 s.

And don't get me started that we don't have any. I've hatched a secret
plan to use that high intensity solar energy at a facility orbiting the sun
inside the orbit of mercury to power my accelerators. Sure it may
  take my 50 years or so to make enough to launch a mission but, what
the hey, that's a lot better than traveling the slow way. You have to
take the long view. (Oh yeah the crew will be stored eggs and revived,
raised, and educated on site by robots to save on mass)

Shhh. Don't tell anyone.

Yeah, and you're an idiot. The biggest reason the French suck is because they had the right, the obligation, the duty, to prevent the Germans from building up their military after WWI, and they gave in, they failed, they surrendered. Every death in WWII, including millions of Jews and Russians, can be laid at the feet of the French for failing to live up to the terms of the Treaty of Versailles which gave them the right to march into Germany at any point they desired and smack them down. But the french were afraid, and thus were key to making WWII happen by their inaction.

Then, of course, there's this fact: "There has rarely been a war in which France did not surrender or win with the help of making allies with another country that would do all the work for them."
http://www.personal.psu.edu/users/u/u/uup101/France.htm

The meme won't die because it's based in historical fact. Sure, France helped the US out mightily during the Revolution--but not due to any great purpose or charity, only to hurt the British. And there's no one more smug, more racist, that I know of than the French.

The french are wussy surrender-monkeys and smug bastards. I'm sure it won't be long now before the french give in to the muslim elements in their society and become the first muslim european state.

For a wonderful introduction to peer review, you could do worse than read this:

http://www.cgoakley.demon.co.uk/qft/corres.pdf It is an exchange, carried out over several years, between a man who believes he has solved quantum field theory, and the reviewers who carefully look through his papers when he tries to publish. They come up with good points and ways to improve the paper, but he resubmits and resubmits until he finds somewhere that accepts it. Along the way, he gets increasingly rude and angry, while the reviewers remain polite and engage carefully with him.

My favourite part is that it's published on the guy's personal website, although he really doesn't come out of it well.

Some of the reviewers don't seem too polite:

In spite of its pompous language, this paper is, in fact, less rigorous, from the mathematical point of view, than ordinary perturbation theory which can be found in any good textbook or review article.

No relevant problems of contemporary QFT are considered in this paper. I recommend its rejection.

Anyone who has attempted to get a scientific paper published has run into these guys - rude, opinionated, and lazy reviewers. It's not just cranks who get these responses, either.

Find a way to measure relative productivity, and relative error rates, for before and after you had to stop using music.

Use objective facts to show your boss what a twat he is.

Why do your own half-assed study with way too few participants when others have already done better ones? See, e.g. Fox & Embrey Music - an aid to productivity Applied Ergonomics 1972;3(4):202-205, or perhaps better still Lesiuk The effect of music listening on work performance Psychology of Music 33(2):173-191 (available as pdf here). Findings from abstract:

Results indicated that state positive affect and quality-of-work
were lowest with no music, while time-on-task was longest when music was
removed. Narrative responses revealed the value of music listening for positive
mood change and enhanced perception on design while working. Evidence is
provided of the presence of a learning curve in the use of music for positive mood
alteration. Overall, the study contributes to the development of a model that
aspires to elucidate music and workplace interactions; as well, it has implications
for organizational practice.

LOL, your comment reminded me of this old Apple commercial (needs Quicktime): http://pulsar.esm.psu.edu/Faculty/Gray/graphics/movies/fullsupport.mov

I thought about moderating this poster up, but decided that there's a fair chance I'll participate, so I'll back up this.

I do research in programming languages (almost more "program analysis" at this point). The two top conferences in our area are POPL (Principles of Programming Languages) and PLDI (Programming Language Design and Implementation). At least in my area, MSR (Microsoft Research) publishes at least on par with a top-tier research university, and judging by the program for POPL 2010, even more so.

Between MSR Redmond, MSR Bangalore, and MSR Cambridge, there are MSR people who are coauthors on eight of the 41 papers accepted. The next runner up: Cambridge University, with authors on "only" 3 papers. Most of those papers are collaborations between MSR people and professors at other universities. Even so, two papers are co-authored entirely by MSR employees. That ties Cambridge, Harvard, Tohoku University, UCSD, and U Penn.

MSR has also done much of the work over the last decade that's been pretty groundbreaking in software model checking (the SLAM work, and more recently, Yogi).

I can't speak to how dominant they are in other subfields, but they're at least well-respected all-around.

The two engines gave quite similar results for such a clear unambiguous and uncommon term. This implies they are spidering with similar coverage.

Most of the results that are controversial/missed by a search engine will occur at the periphery of the web. There's one giant honkin' connected in the Web, so it's quite unsurprising that the first few dozens of results for a well-defined, common term will show up in all major search engines. Your search does NOT imply that they are spidering with similar coverage. See the paper by Broder: Broder et al

Google's results were clearly more relevant. This implies that Bing's ranking algorithm is still not as good as Google's.

Anecdotes and subjective judgments do not result in implications, unless you're Sarah Palin.

one occurs overwhelmingly in an unwanted context. Bing borks them.

This is interesting, but again, "miserable failure" can be used as an anecdote for Google "borking" too.

Sorry about that.

The original MBH98 "Hockey Stick Paper" is "Global-scale temperature patterns and climate forcing over the past six centuries" by Michael E. Mann, Raymond S. Bradley & Malcolm K. Hughes from Nature 392, 779-787 (23 April 1998) | doi:10.1038/33859

The latest work, correcting for the concerns brought up in 2003 is "Proxy-based reconstructions of hemispheric and global surface temperature variations over the past two millennia" (Mann et al. Proceedings of the National Academy of Science. USA 105, pgs 13252–13257; 2008)

PNAS is open access, and the Nature article linked to in this comment comes from Mann's homepage, so you should be able to read them both. My original link to MBH98 was to Nature's site, and they want you to pay $32 to read the article if you don't already have a license.

Until recently, I was a vocal opponent of PowerPoint. I had read Tufte's essay and applied the assertion-evidence structure to my slides. When presenting certain types of data to an english audience, these measures are effective.
But when a relevant percentage of the audience does not understand English, or when the presenter does not speak English, writing the entire presentation down on the slides and reading off the slides is a more effective way of communicating. ESL students are more able to comprehend what they read than what they hear. What 'using powerpoint well' means is a function of the audience and the material.

Also just found another one, but this is a bit heavy on the maths:

Network information flow - R Ahlswede, N Cai, SYR Li, RW Yeung, 2000
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.89.4568&rep=rep1&type=pdf

I think this was the original paper that pointed out that it's possible to exceed the naive "optimal broadcast" efficiency of a single source on a switched network, by allowing intermediate nodes to perform some computation.

Mind you, this is only tangentially related to current P2P systems, but it's still interesting.

Since when did code optimized for speed go out of fashion?

Since 1974, when Donald Knuth claimed "We should forget about small efficiencies, say about 97% of the time: premature optimization is the root of all evil."

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.103.6084&rep=rep1&type=pdf

Of course there's a balance to be made - don't write stuff that's blatantly inefficient. But neither should you spend a week optimising a routine, when a $1000 worth of hardware removes the problem.