Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Comment Re:Leave. (Score 1) 433

There is no way to trace what they did, no way to confirm their methods. Sadly the masses are not equipped to scrutinize the nonsense. [Steve A Morris, 2017-01-11]

You can trace what Hausfather et al. 2017 did by downloading the code they made freely available at bit.ly/2jXSy7G. You can confirm their methods by reading the full paper and following the links at the end which lead to all the data they used. Interested members of the public can read or watch the background they shared.

... they simply don't use 1/3 of the ARGO datasets because its data is "more ambiguous". Translation: "It doesn't fit our needs." [Lonny Eachus, 2017-01-11]

Read the paper to see if Lonny's "translation" is reasonable: "... Two of the three Argo near-SST records assessed, APDRC and H2008, agree well with the buoy-only and satellite-based records and suggest a cool bias in ERSSTv3b during the 2005-2015 period, when sufficient Argo data are available (Fig. 3). The RG2009 series is more ambiguous, with trends that are not significantly different (P > 0.05) from either ERSSTv3b or ERSSTv4. ..."

Lonny Eachus is wrong to claim that Hausfather et al. "simply don't use 1/3 of the ARGO datasets" (presumably a reference to RG2009). They used 3 independent Argo near-SST (near sea surface temperature) datasets, and reported the results from all 3 datasets. Anyone who reads the full paper will see that they mention RG2009 a total of 17 times while reporting the results of using that dataset.

... the study's argument is rather weak. ARGO data has best coverage, best instruments. Yet they arbitrarily throw out 1/3 of the ARGO data sets because they don't agree with their preconceptions. ... In sum, it appears that this paper committed the same likely error as Karl et al. That is to say: ignoring arguably better data because it doesn't fit their preconceptions. [Lonny Eachus, 2017-01-11]

Wrong. Hausfather et al. didn't "throw out" or "ignore" 1/3 of the Argo datasets. Look at figure 3 (backup). They show the results of all three Argo datasets, including four instances using the RG2009 dataset which Lonny baselessly accuses them of "arbitrarily throwing out" and "ignoring".

Paper: (1) "We constructed our own data set from other data sets." (2) Oops. But we left some out. "(3) "We find MOST of the data we used does not match our new contrived data set. So we will ignore it." [Lonny Eachus, 2017-01-11]

Again, Hausfather et al. didn't "leave out" or "ignore" the RG2009 dataset. Look at figure 4 (backup). They show the results of all 3 Argo datasets, including the RG2009 dataset which Lonny baselessly accuses them of "ignoring".

Figure 4 examines four composite SST records: ERSSTv4, ERSSTv3b, HadSST3, and COBE-SST. These composite SST records are compared to instrumentally homogenous datasets (which just means "from a single type of instrument"): buoys, CCI (satellite), and all three Argo near-SST datasets. Figure 4 subtracts all those datasets from each composite SST record, then calculates the trend. If a differenced trend includes "zero" inside its 95% confidence interval, scientists say that particular instrument's trend agrees with that particular composite SST record's trend at the 95% confidence level.

For both examined timespans, buoys and CCI agree with ERSSTv4 at the 95% confidence level, and disagree with all the other composite SST records. The H2008 Argo dataset disagrees with all composite SST records because it shows more warming than all composite SST records, although ERSSTv4 is the closest match. The APDRC Argo dataset agrees with ERSSTv4 and COBE-SST. The RG2009 Argo dataset (which Lonny wrongly claims they "ignored") is in fact the last dataset shown in figure 4. RG2009 agrees with all four composite SST records at the 95% confidence level. That's what Hausfather et al. meant when they said RG2009 is "more ambiguous".

Paper: (1) "We constructed our own data set from other data sets." (2) Oops. But we left some out. "(3) "We find MOST of the data we used does not match our new contrived data set. So we will ignore it." [Lonny Eachus, 2017-01-11]

Presumably Lonny's "new contrived data set" is ERSSTv4, which Jane/Lonny has complained about ad nauseam. Look at figure 4 again. Buoys and CCI satellite datasets agree with ERSSTv4. The Argo APDRC and RG2009 datasets agree with ERSSTv4, but they also agree with other composite SST records so those results are more ambiguous. The Argo H2008 dataset disagrees with all composite SST records because H2008 shows more warming than all of them including ERSSTv4, though ERSSTv4 is the best match.

In other words, they didn't ignore any data, and most of the data matches ERSSTv4. In fact, the RG2009 dataset which Lonny wrongly claims they "simply don't use" is "more ambiguous" precisely because it does match ERSSTv4 (and all the other tested composite SST records). The Argo H2008 dataset is the only one which doesn't have a trend matching ERSSTv4 at the 95% confidence level (because H2008 shows more warming than ERSSTv4) and it also shows that ERSSTv4 is a closer match than any other tested composite SST record.

That really is what they did, though. As I described. They transformed data in ways that are not 100% clear. [Lonny Eachus, 2017-01-11]

No, what Lonny described really isn't what Hausfather et al. 2017 did. In fact, it's hard to imagine how Lonny's description could have been more wrong. See above. Or just read the paper to see that they didn't ignore data and made their methodology 100% clear by making their code freely available.

I find it amusing that climate scientists are prone to argue "surface temp. records are better than satellite". But then claim that the satellite record is "better" than ARGO floats. Just as they did with satellites, climate scientists crowed about their new ARGO floats. "Best thing ever." Then, just like satellites, when the new data does not fit their preconceptions, they just omit it. That has been a very noticeable pattern in the field of climate science. ... a pattern of behavior is a pattern of behavior. Always excuses to omit inconvenient data. [Lonny Eachus, 2017-01-11]

Again, Lonny's delusional narrative where "scientists crowed" about satellite data before "omitting" them is completely baseless. And even though he probably won't ever admit it, deep down Lonny should realize that his new delusional narrative about Hausfather et al. "omitting" data was also just shown to be completely baseless.

Again, Lonny's just projecting. Jane/Lonny previously cited ocean heat content (OHC) measurements based on Argo and satellite data until I showed him that those data doesn't support his incorrect claims that there hasn't been any global warming for 18 years. For years, Lonny has shown a pattern of behavior where he ignores the "best measure" of global warming: OHC data from Argo which reveal ~90% of Earth's added heat. Maybe Lonny omits those data because they don't fit his preconceptions?

Comment Re:Leave. (Score 1) 433

Regarding recent Hausfather et al. paper, which is the source of the latest hype about "no pause": As Anthony Watts points out, the study only goes to 2015, and the middle of its strong El Nino. If it had gone to the present, after record cooling, it would show less or no overall warming. Quote Watts: "Personally, it looks like ignoring the most current data available for 2016, which has been cooling compared to 2015, invalidates the claim right out of the gate" Here's the quote and some other criticisms of Hausfather et al. dailycaller.com/2017/01/05/new-study... [Lonny Eachus, 2017-01-11]

No. When Hausfather et al. 2017 was published (long after it was submitted) the most current available NOAA data ended in November 2016. Nick Stokes showed that even if Hausfather et al. had used a time machine to include those data when submitting their paper, it would have showed more warming. Even the silly opinion piece Lonny linked notes that "climate models will more closely match observations once 2016 data is included".

... its conclusions might have been different after the record cooling we've seen, post- El Nino. [Lonny Eachus, 2017-01-11]

Ironically, Zeke Hausfather showed that including all the 2016 data available at publication actually increases the observed warming trends compared to their paper's conclusions using data through 2015. This is still true using the full 2016 NOAA data which just became available on January 18. Lonny could verify this by repeating these least squares trend estimates with the monthly data, or just noticing that the annual ocean average was even higher in 2016 than in 2015. Zeke Hausfather challenged Anthony Watts to find an ocean temperature record that was cooler on average in 2016 than in 2015. Watts couldn't name one or bring himself to retract his claim. Can Lonny?

... Personally, it looks like ignoring the most current data available for 2016, which has been cooling compared to 2015, invalidates the claim right out of the gate. ... the data only goes to December 2015. They've missed an ENTIRE YEAR's worth of data... Looks like a clear case of cherry picking to me, by not using all the available data. ... [Anthony Watts, 2017-01-04]

Watts accuses Hausfather et al. of ignoring the most current data and missing an ENTIRE YEAR's worth of data. Since Hausfather et al. 2017 was submitted in early 2016, they'd have needed a time machine to include the ENTIRE YEAR's worth of data that Watts accused them of ignoring and missing. In contrast, Sou notes that Anthony Watts presented an AGU poster in 2015 without data from 2015, 2014, 2013, 2012, 2011, 2010, or 2009.

Global Warming Lies: thinkprogress.org/climate... "It's Happening Now"
No, it isn't. NO warming — zero — for 18 years! [Lonny Eachus, 2014-11-06]

Nonsense. The Earth continues to warm. I'd already told Jane/Lonny Eachus that his very ironic accusation is wrong. I'd even shared code and data showing that there hasn't been a statistically significant change in the warming rate, and there isn't a statistically significant difference between the projected and observed trends.

But more importantly, I'd also already told Jane/Lonny that ~90% of Earth's added heat goes into ocean heat content (OHC) measured by Argo, etc. Why are Lonny Eachus and "FriendsOScience" ignoring the fact that OHC keeps rising when they keep wrongly claiming that there hasn't been any global warming for 18 years? Are they ignoring Argo data because Argo data show that Earth continues to warm, which doesn't agree with their preconceptions?

Comment Re:Hmm (Score 1) 389

Emphatically YES! Smarts in one narrow field doesn't guarantee smarts in every field: John Podesta is a Smart Guy, but he was stupid enough to fall for a phishing attack.

Yeah, the lesson I took from that is pretty simple and clear: DO NOT READ EMAIL AT 4 AM WHEN YOU ARE NOWHERE NEAR IN CHARGE OF YOUR FULL FACULTIES. Not sure why Podesta hadn't already figured that one out.

Comment Re:Primary factor (Score 1) 455

Let's understand the REAL issue here; the PATENT prevented everyone else from implementing a safety.

(Hyperbolic emphasis removed.)

I presume you mean "safety feature." In that case, you need to also understand that the fellow was using Apple's products and functionality, so their electing to implement or not implement the feature and thus blocking anyone else from developing the same has no bearing. You might also want to look at other instances where one company has patented some feature or product and that did not block other companies from producing highly similar features or products based on alternate implementations. There are many, many instances and examples of such cases. At the same time, we don't know that Apple, electing to not implement the patent, had not been approached to license the IP by manufacturers who wanted to implement the feature on their hardware. The argument you are proposing is that having a patent irrevocably blocks any development along a given line of inquiry, and that assumption is not exactly correct. Rather not, in fact.

Apple is being sued in this case, we might readily presume, because they have deeper pockets than anyone else, certainly deeper than the driver who so unfortunately caused the accident by his negligence to the task at hand of keeping his attention to the road. I suspect the suit will not succeed, and hope that the family can be brought some solace by owning all future wages ever earned by the liable driver.

Comment Re:Legal reference (Score 2) 163

Whether the amount the company is charging is an accurate reflection of their costs, or whether they are able to make a profit at it are irrelevant considerations. Whether the business model is a potentially successful one is not a legal question. And the simple counter-argument is that many, many, many businesses offer below-cost services in order to seed growth, especially early on in their existence. Even mature businesses offer so-called loss-leader specials that are intended to attract customers, even if they are not strictly money-making in an of themselves.

Whether the $1 final cost to the customer is sustainable by VidAngel is irrelevant: they could change their prices tomorrow, and for all we know, already have a price increase path plotted for the future.

Comment Re:Dear Developers... (Score 1) 173

Agreed. The only instance I can think of where a major UI change was definitively for the better was when GIMP tossed that horrible multi-window idiocy for the unified window presentation. That was a clear win (and one that users were clamouring for extensively). Other than that, though, it's all been for-the-worse. The basic menu is a great structure, but what makes it super-duper is having a help system that allows you to search for functionality without having to resort to Google.

Comment Re:Credible study? (Score 3, Informative) 279

you need to measure all mammals or a decent cross section.

Did you read the scientific article? They measured a large number of species, but limited their study to primates and carnivores, as it says in the title of the study:

Postcopulatory sexual selection influences baculum evolution in primates and carnivores

within the scientific article they give the details as to how many:

A supertree phylogeny of 5020 extant mammals was used to reconstruct the ancestral states of baculum presence across the mammalian order.

That sounds like a decent cross-section. Five thousand species.

Secondly, I am not sure how they came to their monogamy theory.

Did you read the scientific article? It's pretty well explained there. They correlated the mating strategy of each species with baculum length. The homo erectus link was done by the Guardian article reporter, though.

Primates in polygamous mating systems were found to have significantly longer bacula than those in other mating systems (n = 65, p = 0.032).

and later

Two more phylogenetic t-tests showed that primates in polygamous mating systems and seasonally breeding primates had significantly longer bacula than primates in other mating systems and those without a seasonal breeding pattern, highlighting the importance of postcopulatory sexual selection as a driver of bacular evolution.

Comment Re:Credible study? (Score 1) 279

chimpanzees do have a baculum, so the correlation is simply not there between longer intercourse and existence of a baculum.

For an otherwise cogent and reasoned posting, you kind of lost it there. Correlation is not causation, or in this case, the correlation may not be perfect since you have identified an exception. Or, perhaps, the correlation is indeed graduated such that the larger the baculum, the longer the intromission (as just a wild-assed guess). Correlation does not need to be 100% in order to observe a valid link.

Indeed, if you read the article's abstract (despite the broken link in the summary), you'd find that the authors are EXACTLY correlating baculum length with intromission duration (emphasis added):

The extreme morphological variability of the baculum across mammals is thought to be the result of sexual selection (particularly, high levels of postcopulatory selection). However, the evolutionary trajectory of the mammalian baculum is little studied and evidence for the adaptive function of the baculum has so far been elusive. Here, we use Markov chain Monte Carlo methods implemented in a Bayesian phylogenetic framework to reconstruct baculum evolution across the mammalian class and investigate the rate of baculum length evolution within the primate order. We then test the effects of testes mass (postcopulatory sexual selection), polygamy, seasonal breeding and intromission duration on the baculum in primates and carnivores. The ancestral mammal did not have a baculum, but both ancestral primates and carnivores did. No relationship was found between testes mass and baculum length in either primates or carnivores. Intromission duration correlated with baculum presence over the course of primate evolution, and prolonged intromission predicts significantly longer bacula in extant primates and carnivores. Both polygamous and seasonal breeding systems predict significantly longer bacula in primates. These results suggest the baculum plays an important role in facilitating reproductive strategies in populations with high levels of postcopulatory sexual selection.

And in case the terminology used here seems odd, "predicts" in this context means they have fitted a model (that's the Markov chain, Bayesian framework stuff) that uses various parameters to understand underlying structure in the noisy data, and that if you vary a given parameter, like baculum length, the model will predict a values for another parameter, say intromission duration, with good correspondence to the actual data, suggesting the model has captured a link of some sort. The key to understanding this is that the data will of necessity be noisy (all biological data are noisy), and the model, if it's a good one, will reject that noise and capture the underlying structure, thus there will likely be deviations from perfection.

For me, educated as an engineer and trained as a biologist, it was a shock to discover that a model in biology was considered to be very good fit when it was only 30% off the mark, whereas in engineering a model was called good when it was only 1% off the mark. Engineering is a far more refined discipline than biology.

Comment Re:Reason for caution: mechanisms not understood (Score 2) 293

... use resources more efficiently ...

A wise person to whom I'm distantly related argued for in front of a small european parliment for that government to put its effort into efficiency thus: energy conservation provides temporary relief that disappears once economic conditions improve, whereas advances in energy efficiency have indefinite payoff.

Comment two stacked LCDs? (Score 3, Insightful) 103

Reading between the lines, it sure sounds like they just stacked two LCDs and bumped the brightness of the light source. Mind you, that's a very good idea. The new underneath layer probably only needs single R/G/B group resolution in order to achieve the claimed specs, making it somewhat easier to manufacture, although alignment is still going to be important to get right, as will appropriately close bonding of the two planes to control leakage from one luminance cell (for want of a better word) to the neighboring RGB cells in the color layer.

A highly-motivated enthusiast might be able to get close to the same results by merging two existing IPS monitors and bumping the light source brightness.

Comment Re:Init alternatives (Score 3, Insightful) 338

The biggest improvement over antique boot systems ...

That there is the heart of the problem, an attitude that anything old is necessarily bad. That your otherwise calm and reasoned presentation allowed this pejorative to slip in belies the psychological bias that underlies the wide arguments on the subject.

Lest we forget, Linux as a whole turned 25 recently. That's antique. Are you giving up the entirety because it's old? Your favorite editor is probably (just based on popularity) is either emacs or vi / vim. They are very, very old (heck, I've been using emacs since the early 1980s!). Are you dumping them because they are old? I hope you see why calling something "antique" is ill-conceived.

Now to make sure that my point is being made clear, allow me to be explicit: old does not necessarily mean bad, but it does not necessarily mean good, either. Things that are old now were once shiny and new, and weren't necessarily an improvement when they were introduced. But change merely for the sake of change -- which seems to be what was behind debacles in KDE, Gnome, systemd, and Wayland to name a handful -- is wasted effort. For systemd in particular, the primary argument for using it seems to be parallel init, something that as many others have pointed out really isn't much of an issue these days since (a) Linux is generally stable enough that reboots are rare (although there are specific use-cases that benefit, like demand-based VM creation), and (b) computers have become generally fast enough that reboots are inherently speedy.

Comment But, why? (Score 1) 243

Yes, sure, an interesting thought experiment, I suppose. Maybe. If you're the sort of psychopath who likes to pull legs off of small insects and animals just to watch them die. And, if that's the case, well, you need to be removed from direct contact with society and should be seeking treatment, possibly including protection from yourself. There is no legitimate use that comes to mind for a USB-killer other than to intentionally destroy property (unlike, say, a firearm which has legitimate uses beyond the raw ability to kill or maim). Moreover, it would seem to be targeted toward public-facing USB ports which are, in general, a public good, and destroying a public good brings us back to the psychopath issue.

For everyone else, well, that sort of creative energy is useful put to more positive efforts.

Comment Re:Can you see Google's Code? (Score 1) 385

Do not ascribe to malice that which can be explained by ineptitude. Or something like that.

My understanding is that the result on Google, both displayed search results and auto-completion suggestions, are based on a large number of factors including (wait for it) your personal past history of typing, and the most popular results.

Assuming that a "voting for ..." search would autocomplete with both main candidates equally assumes that there are as many people typing one candidate as the other. That is an unfounded assumption that is likely false. I searched a few times for a third-party candidate and lo! my autocomplete changed. The horror! Bias! Lynch them! Or, wait, maybe if I were searching for widgets that might be a good idea. Maybe if I were looking for somewhere to vacation, that might be a good idea. Maybe under circumstances that are not so emotionally incendiary, we might want exactly this behavior because it works very well.

Any evaluation that demonstrates bias and is shocked by the results (or merely reports them) must prove unequivocally that the assumption of a lack of fundamental underlying bias in the data exists. A fundamental equality of data (e.g. the same number of web sites for the two candidates, the same number of twitter posts, etc.) is unlikely to be true in any of the cases being discussed in this thread.

In other words, the original posting is not news. It is, to use the current vogue terminology, fake news. Lying with statistics. Click bait. Something to be ignored.

Slashdot Top Deals

/* Halley */ (Halley's comment.)