Slashdot Mirror

placement new doesn't work without nullifying a few things. Automatic cleanup on scope exit doesn't work for locks in the kernel. See below ... Much more ...

placement new/delete are noexcept functions. But, they call std::terminate--not acceptable. The only thing that works is an alloc function that returns NULL (or (void *) -errno). Returning null is not fatal in the kernel. The caller must be able to deal with it (usually returning -ENOMEM). So, the [global] new/delete must be changed. Also, placement delete has problems [I've left off the backslashes for clarity]:

#define GETPTR(_ptr,_typ,_siz)
switch (_typ) {
case 0:
_ptr = alloca(_siz);
break;
case 1:
_ptr = kmalloc(GFP_KERNEL,_siz);
break;
case 2:
_ptr = kmalloc(GFP_ATOMIC,_siz);
break;
case 3:
_ptr = slab_one(_siz);
break;
case 4:
_ptr = slab_two(_siz);
break; // ...
}

void
myfnc(int typ)
{
void *ptr;
GETPTR(ptr,typ,23);
class abc *x = new(ptr) abc(19,37);
}
At this point, a delete operator [even a placement version] has no idea which pool to release to because there's no way to pass typ to it. You might be able to create a contructor abc(typ,19,37) but that adds an extra member element to hold typ so the delete operator can get at it, but that's additional overhead/complexity that C doesn't have. It might be possible to make it work by casting typ to void* and using that as the pointer:
class abc *x = new((void *) typ) abc(19,37);
and have the class specific new operator use GETPTR internally. I tested this and it works. However, I haven't yet been able to get the corresponding placement delete to work as a class specific overload [yet]. In trying to find the way, I came upon:
http://www.scs.stanford.edu/~d...
It's fairly detailed and lays out a [pretty strong] case against using the new operator [more eloquently than I could do here].

A lot of kernel code puts definitions in the usual place [top of function body] for C. In C++, this invokes the constructor, which is not what you want. The reason is that [say] 10 vars are defined. The function does a quick check on args and does a non-standard return -EINVAL. All that wasted create/destroy. This may be harmful if the constructors have side effects such as lock acquisition. Note that doing a [wasteful] lock followed by an immediate unlock [to satisfy having a destructor do lock cleanup] is a non-starter in the kernel [you'll never get such code checked in/signed off on]

So, you'd have to go through every kernel function by hand [there are 16.9 million lines of source code] and move the definitions down:
{
struct foo x;
if (bad_news)
return -EINVAL; ...
}
{
if (bad_news)
return -EINVAL;
struct foo x; ...
}

You can't put a lock release in a destructor because you'd need an extra member var that would have to be set/cleared when you acquire/release a lock. That's because the destructor has to have some way of knowing whether to suppress the lock release. So, you're adding an extra variable [that isn't needed in C] just to prevent an attempt to release a lock that was never acquired in the first place. More overhead and slower [and more complex] than its C counterpart.

I

no machine writing, no photographs where there's not even the slimmest claim of artistic composition or whatever.

This is not true in the US or the UK. As an example, Wolfram Alpha copyrights all its output.
http://blogs.law.stanford.edu/...

It's not just about Zoe Quinn and Anita Sarcasessian. There is a great deal of credible evidence that harassment of women gamers is very common and very ugly.

https://undergrad.stanford.edu...

Here's the Boothe Award-winning paper:

https://stanford.box.com/share...

Stanford says it's an "internal policy": https://cyberlaw.stanford.edu/...

All donors to the Center--and to Stanford more generally--agree to give their funds as unrestricted gifts, for which there is no contractual agreement and no promised products, results, or deliverables.

But this makes absolutely no sense. If all money goes into a general fund, there's no distinguishing "whose" money it is, it's Stanford's money.

And I can totally imagine them coming home and their grandparents asking them "Where did you go this year?" in the most obnoxious wasy possible, like all grandparents do. Good luck explaining this one!

Yeah, yeah. Been there, done that. About 375 years ago.

Back when Vesuvius was actively erupting in the 1630s, I decided to take a closer look. While the volcano was still smoldering and active, I hired a local guide to take me to the top, then was lowered into the crater to take some scientific observations and temperature measurements. You wanna read more about it? Here's some info on my book. (Of course, science has progressed a lot since then.)

We didn't have fancy videocameras back then, so I had to make make my own drawings of what I saw.

Try reading whole sentences. The following is a quote from this article.

Due to evaporation losses from the exposed water surface and mechanical efficiency losses during conversion, only between 70% and 85% of the electrical energy used to pump the water into the elevated reservoir can be regained in this process.

Here is a quote from another article

The cycle is generally about 80% efficient, with losses due to water evaporation and engine non-idealities.

And another article.

Pumps and turbines (often implemented as the same physical unit, actually) can be something like 90% efficient, so the round-trip storage comes at only modest cost.

Please note that 90% pump efficiency + 90% turbine efficiency equals 81% overall efficiency.

Here is another;

First, the charging process in pumped hydro storage is affected by the pump efficiency that pumps the water into the upper reservoir at times of low electrical demand. The losses during discharging process on the other hand are caused by the turbine operation to generate electricity at peak load periods. The total charging and discharging rate is given by calculating the product of the efficiencies of pipe (friction losses) and the mechanical equipments

The hourly evaporation losses is assumed to be negligible because the amount of water evaporated is far too small compared to the total water volume in the reservoir

That paper quotes efficiency at 75 – 85 percent.

Here is an article stating that evaporative losses are minor;

North Eden Creek will be the primary source of water for the initial fill of the lower reservoir. Water rights will need to be secured, both for the initial fill and annual evaporation maintenance. The advantages of this system are that once the initial fill has occurred, the only water needed will be a small amount to offset annual evaporation from the reservoirs. The precipitation and evaporation balance will result in an annual water loss of approximately 0.2m over the total surface area of both reservoirs.

While higher in the desert evaporative losses do not effect cycle efficiency significantly.
Need I go on? Yelling without up backing you statement with references just weakens your case.

An accelerator can disrupt the fission products directly. You are thinking of the transuranics with your spallation target. http://large.stanford.edu/cour... But, the fission products can themselves be proton targets and be disrupted right down to hydrogen.

I understand that you have a strange love for nuclear power. But for those of us who see it realistically, your love of power is a classic of mythology which always ends badly. Nuclear power has its place in naval propulsion, but in a civilian context it is a very poor choice. It is time to clean up your mess.

Join the Quake catcher network of Stanford. You can order a sensor, but your laptop can detect them too.
http://qcn.stanford.edu/

The detected earthquake
http://qcn.stanford.edu/earthq...

Bert

Join the Quake catcher network of Stanford. You can order a sensor, but your laptop can detect them too.
http://qcn.stanford.edu/

The detected earthquake
http://qcn.stanford.edu/earthq...

Bert

I don't know what shit-stained pube-licking school you attended, but institutions like Stanford are pretty fucking conservative. Don't worry though, we know that a brain-dead little AC faglet such as yourself could certainly never understand reality. Please go back to shitting yourself stupid and let the adults have a grown-up conversation.

However, even the idea of individual justice seems ambiguous in regard to scope. Plato in the Republic treats justice as an overarching virtue of individuals (and of societies), meaning that almost every issue he (or we) would regard as ethical comes in under the notion of justice (dikaosoune). But in modern usages justice covers only part of individual morality, and we don't readily think of someone as unjust if they lie or neglect their children--other epithets more readily spring to mind. What individual justice most naturally refers to are moral issues having to do with goods or property. It is, we say, unjust for someone to steal from people or not to give them what he owes them, and it is also unjust if someone called upon to distribute something good (or bad or both) among members of a group uses an arbitrary or unjustified basis for making the distribution (this last aspect of individual justice obviously has reference to social or at least group justice). Discussion of justice as an individual virtue standardly (at least) centers on questions, therefore, about property and other distributable goods.

http://plato.stanford.edu/entr...

Well, this article he wrote is nonsense. I know nothing else about the guy.

He just takes every controversy and paints it as an unsolvable failure of the iron-fisted Wikimedia Foundation.

Funny -- it seemed to me like one of the most insightful and relatively balanced pieces on Wikipedia I've read in some time. Despite having a number of serious complaints, the author also talks about very positive aspects of the event, as in: "Wikimania was in many ways an inspirational event. There was a palpable sense of enjoyment and celebration in the air..." and later in the final conclusion "As I travelled to Wikimania, I worried that I might hate it. But my worst fears did not materialise." He clearly cares greatly about Wikipedia and wants to make it better.

In fact, some of the potential solutions he mentions address the biggest problems of Wikipedia and could finally be the path to solve them. For example:

Medical content, notably the current initiatives to have medical articles peer-reviewed by academic experts (Cancer Research UK is involved, and is now hosting a Wikipedian in Residence), and provide readers with a permanent and prominent link to that peer-reviewed article version. It's an excellent idea that in the long run could also be transferred to other topic areas. Experts might be more inclined to contribute and review articles if their work is guaranteed some lasting presence. We hope the Foundation will support that effort.

Wikipedia has grown over the years by leaps and bounds with all the wonderful contributions from random people. But for articles that have achieved a relatively good status, Wikipedia is spending more and more time fighting the "barbarians at the gates" who want to vandalize, post misleading information pushing an agenda, and just random editors with little expertise who wikilawyer their way into having the article the way they like, regardless of an expert consensus on the topic. All of this could be solved by keeping articles more "stable" (maybe have a separate proposed edits page, or an "experimental" page that could be edited by anyone and is not the default) and incorporating advice from subject matter experts.

The Stanford Encyclopedia of Philosophy, which is older than Wikipedia and the best resource on philosophy on the internet, shows how this can be done well. Wikipedia wouldn't have to let go of the option for the general public to edit pop culture articles on their favorite Star Trek or Buffy or Friends episode or whatever -- but for subject matter where there is a peer-reviewed expert academic literature available and experts who are willing to contribute, why not help make that possible?

Similar policies could solve some of the biographical article problems brought up in the summary -- even just holding proposed edits in a queue for experienced and validated editors to allow them would prevent nonsense such as that mentioned in TFA where a reporter has to complain about: "I have spoken publicly about my children having been born as a result of fertility treatment. And my Wikipedia page, which I didn't even know existed, contained a phrase along the lines of 'he wasn't man enough to impregnate his own wife'," a statement that went unchallenged on Wikipedia for nearly a month. The author (and the reporter complaining here) is right -- there's simply no excuse for that sort of nonsense, particularly when Wikipedia has such a poor track record of figuring out ways for real people to correct factual problems in articles about themselves.

The goal of PrivateCore's product was to encrypt everything that's outside of the CPU core using software techniques. So once you've done an attested boot and gotten your crypto keys in order, from that point on anything outside the CPU socket is done in an encrypted manner (except I/O to the network I guess, but definitely hard disk and data going to the DRAM, etc.) Their important selling point here was that you could protect against cold boot attacks, DMA data dumps, data sniffers on the DRAM lines, etc. They also claim to have a secure hypervisor (preventing cross-VM thievery) because they've stripped it down to its bare bones, but I believe this ended up being a secondary concern.

Anyway, their goal was to have unencrypted data in the caches, but encrypt the data before it leaves the chips and goes out to DRAM. Their page is mostly high-level marketing fluff, so if they were claiming to do more than this, I missed it. The hardware for encrypted DRAM accesses exists in specialized platforms (e.g. the XBox 360) but doesn't currently exist in commodity x86 server parts. As such, a friend and I sat down for an evening a while ago and tried to work out how they would do this without a DRAM controller that did the encryption for you.

Again, their goal is to have decrypted data in the caches, encrypted data in the DRAM. The crypto routines would have to be contained in software. The major difficulty is that the cache does whatever the cache wants, so it's really rather difficult to say "when this data is leaving the cache, call the software crypto routines." There is no good way for the hardware to tell you it's kicking data out of the cache. (There are academic proposals for this kind of information, but nothing currently exists.)

We thought up of a number of solutions and were able to validate our guesses against their patent submission. I will gloss over some of the deeper details (such as methods for reverse engineering the cache's replacement policy).

The shortened version is:
1) Work on Intel cores that have >=30 MB of L3
2) Run a tiny hypervisor that fits into some small amount of memory (let's say 10MB)
3) Mark all data in the system that is not the hypervisor code pages are non-cacheable
4) The hypervisor also has the crypto routines, so all of these non-cacheable pages can now be software encrypted using the hypervisor's routines. The DRAM-resident data is now encrypted.
4a) Because these were marked as non-cacheable data, the hypervisor is still resident in the cache (it was never displaced).
5) Mark some remaining amount of space (let's say 20MB) of physical memory as cacheable. This physical memory currently contains no data at all.
6) When you want to run a program or an OS, have the hypervisor move that program's starting code into the 20-meg-range (decrypt it along the way) and set its virtual pages to point to that physical memory range
7) The program can now run because (at least some of its pages) are decrypted. They are also cacheable, so it will hit in the cache
8) When you try to access code or data that is still encrypted, it will cause a page fault
9) The hypervisor's page fault handler will get that encrypted data, decrypt it, and put it somewhere in the 20-meg-range
9a) If the 20 meg page is already full of decrypted data, you will have to re-encrypt some of it and spill it back to DRAM (like paging it out to disk).

Because you are only touching ~30 megs of physical memory that is marked as cacheable, you will "never" spill decrypted data to the DRAM. Essentially, they built a system that has 30 megs of main memory (that 30 megs is SRAM in the core), and DRAM is treated like disk/swap in a demand-paging system.

The reason I am convinced this is likely an acquisition-hire

The goal of PrivateCore's product was to encrypt everything that's outside of the CPU core using software techniques. So once you've done an attested boot and gotten your crypto keys in order, from that point on anything outside the CPU socket is done in an encrypted manner (except I/O to the network I guess, but definitely hard disk and data going to the DRAM, etc.) Their important selling point here was that you could protect against cold boot attacks, DMA data dumps, data sniffers on the DRAM lines, etc. They also claim to have a secure hypervisor (preventing cross-VM thievery) because they've stripped it down to its bare bones, but I believe this ended up being a secondary concern.

Anyway, their goal was to have unencrypted data in the caches, but encrypt the data before it leaves the chips and goes out to DRAM. Their page is mostly high-level marketing fluff, so if they were claiming to do more than this, I missed it. The hardware for encrypted DRAM accesses exists in specialized platforms (e.g. the XBox 360) but doesn't currently exist in commodity x86 server parts. As such, a friend and I sat down for an evening a while ago and tried to work out how they would do this without a DRAM controller that did the encryption for you.

Again, their goal is to have decrypted data in the caches, encrypted data in the DRAM. The crypto routines would have to be contained in software. The major difficulty is that the cache does whatever the cache wants, so it's really rather difficult to say "when this data is leaving the cache, call the software crypto routines." There is no good way for the hardware to tell you it's kicking data out of the cache. (There are academic proposals for this kind of information, but nothing currently exists.)

We thought up of a number of solutions and were able to validate our guesses against their patent submission. I will gloss over some of the deeper details (such as methods for reverse engineering the cache's replacement policy).

The shortened version is:
1) Work on Intel cores that have >=30 MB of L3
2) Run a tiny hypervisor that fits into some small amount of memory (let's say 10MB)
3) Mark all data in the system that is not the hypervisor code pages are non-cacheable
4) The hypervisor also has the crypto routines, so all of these non-cacheable pages can now be software encrypted using the hypervisor's routines. The DRAM-resident data is now encrypted.
4a) Because these were marked as non-cacheable data, the hypervisor is still resident in the cache (it was never displaced).
5) Mark some remaining amount of space (let's say 20MB) of physical memory as cacheable. This physical memory currently contains no data at all.
6) When you want to run a program or an OS, have the hypervisor move that program's starting code into the 20-meg-range (decrypt it along the way) and set its virtual pages to point to that physical memory range
7) The program can now run because (at least some of its pages) are decrypted. They are also cacheable, so it will hit in the cache
8) When you try to access code or data that is still encrypted, it will cause a page fault
9) The hypervisor's page fault handler will get that encrypted data, decrypt it, and put it somewhere in the 20-meg-range
9a) If the 20 meg page is already full of decrypted data, you will have to re-encrypt some of it and spill it back to DRAM (like paging it out to disk).

Because you are only touching ~30 megs of physical memory that is marked as cacheable, you will "never" spill decrypted data to the DRAM. Essentially, they built a system that has 30 megs of main memory (that 30 megs is SRAM in the core), and DRAM is treated like disk/swap in a demand-paging system.

The reason I am convinced this is likely an acquisition-hire

Since 'the incident' the police is knocking on doors of young couples living in the Fukushima area and in the fall out zones north east of it, telling the couples: " you know, you should consider to have no children" (Or move away to the far south or Hokkaido)

Can you actually show this, or is this just the latest of the tall tales making its rounds on the anti-nuclear blogosphere? And anyway, even if it did happen in some form, all it would show is that people are afraid and giving each other potentially poor advice. It doesn't show that they're at actual substantial risk of harm, otherwise you could go around telling everybody to stay indoors to prevent them from being run over by cars (you know, this we can actually show to happen).

In Chernobyl the death toll over all is estimated to be a million, roughly. /. posters claim it was 3 or 5 ...

Ugh, not that rag again. Yablokov's publication is a book, not a peer-reviewed scientific paper. It contains tons of errors and was translated and pushed onto the New York Academy of Sciences by known anti-nuclear crazies who aren't above outright falsehoods (like their assertions that Fukushima killed 15000 people in US in the initial 14 weeks after the accident, even though their data is trivially shown to have been manipulated and utterly bogus; Mangano is often seen publishing together with another crazie, Sherman, and they've even been torn a new one by an avid linear-no-threshold-supporting researcher). The Yablokov publication has since been criticized by the NYAS and they've distanced themselves from it. The short story is that the NYAS' reputation was co-opted as a vehicle to fluff up the credibility of an utterly bogus piece of non-scientific writing by anti-nuclear activists.

I witnessed 1986 about a few ten thousand ... it was news every day on TV. I really wonder how people in our days with straight face claim only a few people died.

Oh my, so if something's on TV, it is truth! Well fire the scientists then, obviously all we need to do to determine fact from fiction is to listen to the daily news cycle. Fox News will be pleased.

Luckily the initial disaster in Fukushima was far away from this. However the long term issues we only will know in 30 years ... plus.

Even assuming the fairly uncontended (mainly in anti-nuke cycles) linear-no-threshold dose response model, according to actual peer-reviewed studies, on average we'd expect ~250 excess deaths over the years with an upper bound of ~2500 (and that's assuming no evacuations). Was the accident harmless? Certainly not. Should TEPCO be made to compensate people for their troubles? Absolutely! But this fear mongering using junk science is in no way different to global-warming deniers and 9/11 truthers simply ignoring scientific facts to meet their political agendas. Do be like them.

Here's a good, reliable page on Ebola, Reston variant: (this assumes you don't think Stanford is a cheesy school, or in on the vast conspiracy to supress all conspiracy theories, or whatever).
http://virus.stanford.edu/filo...

from this page
"twelve of the 186 people tested had serological evidence of infection with EBO-R. 22% of the workers at Ferlite Farms had positive IFAT (indirect fluorescent antibody test) titers, which was significantly higher than at the other three export facilities."
              Those infection mumbers are low for a virus that normally attacks humans, like Ebola Marburg, in a setting with no precautions at all and lots of hosts, but the fact that humans have no significant symptoms from it says that the Reston variant virus does not colonize humans at all well, and so are at least marginal support for it being exceptionally likely to survive in the environment, compared to the more human lethal types. This just might indicate that Reston is airborne, but probably just indicates it survives a bit longer on surfaces or takes a little more exposure to some disinfectants to destroy than the commoner Ebola virus types. So you're halfway right about that - Reston is not presumed to have become air vectorable, it's just been raised as a possibility in discussion, and is still rated as less likely than some alternatives.

this particular shipment of nonhuman primates had a far larger number of deaths in Room F than would normally have been expected.
              And there goes your record - Reston is deadly to simians, at least to cynomolgus macaques. Unless you want to stand on your obvious spelling error (yeah, it doesn't kill "semians" - I hear not even Kryptonite kills them), the poster you are "correcting" was correct.

      Given a 25% accuracy rating and four spelling errors and two grammer errors in four sentences you would have a hard time persuading people to reject the conspiracy theory that Donald Trump's hairpiece is an Venusion Brainslug invader.

Are you aware of the scientific papers Folding has published?

Google HQ is in California, so they started there. They've expanded to include Nevada, Michigan and Florida, so far.

Nevada was first to pass the law. This is a good site on the issue of legality: http://cyberlaw.stanford.edu/w...

Right now, Google has special dispensation from the state to drive the Lexus SUVs. Anyone who has spent significant time on Bay Area freeways has probably seen them. But these are more or less research vehicles piloted by engineers. Deployment of a production fleet of driverless Google cars in California may be a different matter, as the state has to come up with regulations to allow for this, and that may for at least another year.

Since legalizing it, Nevada has amended their laws in such a way that makes deployment of the Google car easy. Hence all the speculation and rumor that Google may be deploying in Vegas around the next CES/LV Auto Show conventions.

Folding@Home

SETI@Home

These guys can ALWAYS use more cycles.

The Stanford Linear Accelerator Center (SLAC) also extensively used the Amiga for data visualization (Hippograph), capture, and documention (TeX) in the late 80s and just into the early 90s. A few great (mythical?) Amiga applications are hosted there, still:

https://www.slac.stanford.edu/...

http://science.slashdot.org/co...

The content is delivered through RSS. You only need to subscribe to the feeds to get the content. Stanford on iTunes U is a great example.

You can argue if you like that a ~ 27.3% increase is large but I disagree, since climate sensitivity to CO2... [Jane Q. Public, 2014-07-07]

Ocean acidification is independent of climate sensitivity, and it's another reason to be concerned about the unprecedented rapidity of our CO2 emissions.

I would also like to point out again that even if acidification is happening, the RESULTS of that acidification are probably less than alarmists have claimed. Example (2010 article): http://www.rationaloptimist.co... [Jane Q. Public, 2014-06-10]

Lonny Eachus also linked to that misinformation from Matt Ridley, a journalist with a long history of distorting climate science.

In contrast, I quoted from Honisch et al. 2012 (PDF), Knoll et al. 2007 (PDF), and Ken Caldeira’s 2012 AGU lecture. That last link was from my videos section which also includes:

• Andrew Dickson gave a technical 2009 presentation called “Acidic Oceans: Why Should We Care?”
• A series of panels at the 2011 AGU discussed declining reef health and tipping points.

I'm not a chemist or a marine biologist/ecologist, so I read peer-reviewed papers and go to conferences like the AGU to watch lectures by scientists who do specialize and publish in those fields. For instance, consider that 2011 AGU panel on declining reef health. Nina Keul observed one species of foramanifera Glas et al. 2012 (PDF) growing faster as carbonate ion concentration decreases (which happens when CO2 increases). She provided context by noting that this is one species from one experiment, noting that this is like looking at one puzzle piece of a big puzzle.

Then Adina Paytan provides further context by noting that most species aren't like this. She shows Fig. 2 from Crook et al. 2012 (PDF) which shows that only ~3 out of 9 species of coral are present in locations with naturally low pH and notes that "Because these three species are rarely major contributors to Caribbean reef framework, these data may indicate that today’s more complex frame-building species may be replaced by smaller, possibly patchy, colonies of only a few species along the Mesoamerican Barrier Reef."

Finally, Robert Ridin

I'd add: https://en.wikipedia.org/wiki/...
See: http://web.stanford.edu/dept/S...

As for those still alive, Ivan Sutherland:
https://www.youtube.com/watch?...

When asked, "How could you possibly have done the first interactive graphics program, the first non-procedural programming language, the first object oriented software system, all in one year?" Ivan replied: "Well, I didn't know it was hard."

See also: https://en.wikipedia.org/wiki/...
Not all are programmers of course.

But I suppose if you mean programmers who are great closer to the coding skills level and not the "great ideas" level then perhaps it'll be Linus, Michael Abrash and so on.

Coding skills are important but without vision, good ideas and taste you end up with stuff like Desktop Linux or Windows 8. You can have a team of very skillful programmers but if nobody steers the team in the right directions, you won't make much progress.

"97 percent of climate scientists believe human activities are causing global warming.

Actually, it potentially shows some science related to the topic of this article: It could easily mean that the evidence is overwhelming but a minority of 3% is unable to accept it for irrational reasons (it certainly works through the religious pathway with creationists, for example).

When we push beliefs on people and claim it's science, we shouldn't be surprised when they treat it as flexibly as they do any other belief.

All science is automatically a belief. Any piece of knowledge is automatically a belief, as regarded by cognitive psychology. This statement of yours is useless.

Ah, that good old marxist crap about eschewing "formal liberty" for "real liberty". It makes more sense to distinguish between positive and negative definitions of freedom.

You'd have to be a fool to think any one can prosper in this world, that we all make daily, without exploiting others.

Actually, you'd have to be a fool to believe that mass prosperity can ever be achieved and sustained by taking forcefully from others. Quite the other way around, and quite the opposite of your premise.

The number of male and female MD graduates is almost equal: http://kff.org/other/state-ind... and women outnumber men in obtaining a degree: https://collegepuzzle.stanford...

There is no freely-available resource anywhere that can match Wikipedia for accuracy and reliability.

Ah, "freely-available" -- changing the rules, now, are we? Before it was "readily available."

Anyhow, sure there is -- if you're willing to restrict stuff to smaller subsets of knowledge. For example, the Stanford Encyclopedia of Philosophy is, hands-down, the BEST encyclopedic resource on topics in philosophy. It's free, and it predates Wikipedia by quite a few years.

It's much more accurate and reliable than equivalent articles on Wikipedia, because it's stable and overseen by experts, and the articles are frequently updated to take into account new scholarship. It obviously required a lot of organization and funding to get going, and that wouldn't have worked for a lot of Wikipedia -- but there's no reason to keep fumbling in the dark when we could establish something more stable. Even if small sections of Wikipedia were "stablized" a bit, that would be a significant improvement.

If you're looking for a free resource that can match Wikipedia in accuracy, reliability, and comprehensiveness, then... well, you're right. But trivially so, because Wikipedia currently has a near-monopoly in terms of effort into building such databases of knowledge.

Those are terrible counterexamples, because US investments in Europe, South Korea and Japan easily payed themselves back a thousand fold. The cold war was really a form of modern mercantilism. Whereas 18th century mercantilist empires took raw materials from their dependent nations and sent back manufactured goods, 20th century mercantilists (the US, and to a lesser extent the USSR) built silos abroad and sold arms and bonds to their dependent nations. In return the US got enormous shares of stock in companies like Renault, Dassault, Volkswagen, Daimler, Samsung, and Nippon, sources of cheap manufactured goods, and Iranian oil (Saudi oil after the Shah was overthrown).

We Americans like to pretend that we have the largest economy in the world because our parents and grandparents were harder working, more intelligent, and more creative than foreigners. The reality is that we are on top because we were the only nation to come out of the second world war unscathed (thanks you, Atlantic Ocean), and we used that position to take advantage of everyone else.

Winning wars = winning money. Fighting 13+ year unwinnable wars = losing money, but that is a separate issue.

What I'm saying is that while it may be fun to trot out things like the "640K should be enough for everyone" to bust on Bill Gates that is an urban myth and he never said it. Instead, bust on him for things that he *did* do (like hire someone else to pirate CPM). Same for Apple and Jobs (I just have a somewhat better memory for the Microsoft end of things, hence using MS-centric example).

The thing that is funny about the old days of Apple, is that people misrepresent what happened back then too. Everyone claims that Apple ripped off Xerox when they started working on GUI with Lisa / Mac, when one of the alumni of Xerox PARC (and a member of the original Macintosh team) says otherwise. Oh, and Lisa and Mac were already specified to be graphical bitmapped systems before Xerox allowed the Apple team to come in not once, but twice, mostly allowed because Xerox had given Apple VC money.

Never mind that there is a massive gulf between research, and product development. Best example from this particular topic: the original Apple mouse. It went from being a tricky, finicky, expensive piece of lab equipment into a cheap, mass produced, reliable, and easy to use piece of equipment that everyone just accepts as always being there under Apple's development. More info including interviews, design sketches, and documentation here.

For some reason, everyone likes to put research of an idea without actually turning it into something useful on a mile-high pedestal, but turning that idea into something that people can actually use to accomplish things doesn't mean shit. Research is important, but so is development of that research into a useful thing.

What I'm saying is that while it may be fun to trot out things like the "640K should be enough for everyone" to bust on Bill Gates that is an urban myth and he never said it. Instead, bust on him for things that he *did* do (like hire someone else to pirate CPM). Same for Apple and Jobs (I just have a somewhat better memory for the Microsoft end of things, hence using MS-centric example).

The thing that is funny about the old days of Apple, is that people misrepresent what happened back then too. Everyone claims that Apple ripped off Xerox when they started working on GUI with Lisa / Mac, when one of the alumni of Xerox PARC (and a member of the original Macintosh team) says otherwise. Oh, and Lisa and Mac were already specified to be graphical bitmapped systems before Xerox allowed the Apple team to come in not once, but twice, mostly allowed because Xerox had given Apple VC money.

Never mind that there is a massive gulf between research, and product development. Best example from this particular topic: the original Apple mouse. It went from being a tricky, finicky, expensive piece of lab equipment into a cheap, mass produced, reliable, and easy to use piece of equipment that everyone just accepts as always being there under Apple's development. More info including interviews, design sketches, and documentation here.

For some reason, everyone likes to put research of an idea without actually turning it into something useful on a mile-high pedestal, but turning that idea into something that people can actually use to accomplish things doesn't mean shit. Research is important, but so is development of that research into a useful thing.

The test created by Turing specified that there were two subjects that the judge were interacting with. One human and one computer. There is no "not sure" choice.

Yes, you are correct.

It seems that 50% is the correct percentage to me!

Turing's original discussion included the following claim:

I believe that in about fifty years' time it will be possible to programme computers, with a storage capacity of about 10^9, to make them play the imitation game so well that an average interrogator will not have more than 70 percent chance of making the right identification after five minutes of questioning.

This wasn't part of the "test" per se, but where Turing originally thought technology would be 50 years after he wrote those words. (He wrote them in 1950, so that would be a claim about 2000.)

But I believe that's where all this "30%" stuff comes from.

"Ocean Acidification” is an up-and-coming buzz phrase used by global warming alarmists. They say it will harm sea life like coral. [Lonny Eachus]

Caused by CO2, of course. The problem with that theory is that coral evolved when CO2 concentration was *70 TIMES* what it is now. [Lonny Eachus]

"Warmists" like to scare over things like death of coral due to ocean acidification from CO2. Coral evolved at a time of 70x today’s CO2. [Lonny Eachus]

The degree of "doomedness" is highly questionable. I don't dispute that human activities have harmed coral in many cases. But coral evolved when it was both warmer than it is now, AND the concentration of CO2 was many times what it is today. ... [Jane Q. Public]

If atmospheric CO2 increases slowly, ocean pH doesn't change significantly because it's buffered by carbonates and land weathering on long time scales. See Fig. 2 in Honisch et al. 2012 (PDF):

"When CO2 dissolves in seawater, it reacts with water to form carbonic acid, which then dissociates to bicarbonate, carbonate, and hydrogen ions. The higher concentration of hydrogen ions makes seawater acidic, but this process is buffered on long time scales by the interplay of seawater, seafloor carbonate sediments, and weathering on land."

It's incredibly ironic that Jane Q. Public and Lonny Eachus both point to paleoclimate evidence to support their dismissal of ocean acidification. Honisch et al. 2012 also discusses the observed consequences of releasing CO2 more quickly, such as during the end-Permian and PETM.

Paleoclimate evidence shows that ocean acidification depends on the rate of CO2 emissions, not the amount in the atmosphere.

Further, it has been shown that DAILY VARIATION of ocean pH at a given location is greater than any change attributable to CO2. [Lonny Eachus]

Also, studies have shown that the pH in a given location of the ocean typically varies every day far more than any amount that can be attributed to CO2. [Jane Q. Public]

Daily variations can be ~10C or more, but during the end-Permian a ~10C rise in the long term global average temperature coincidentally happened when ~90% of all species went extinct. Furthermore, the marine extinction pattern has ocean acidification's fingerprints on it. Knoll et al. 2007 (PDF) showed that during the end-Permian extinction, ~85% of genuses like coral with aragonite (CaCO3) skeletons went extinct, but only ~5% of genuses like fish with other skeletons went extinct. The rapid CO2 increase during the PETM also led to a similar albeit less severe marine extinction pattern. Again by coincidence?

Corals evolved during the Cambrian Era with CO2 7-20X higher than today. "Ocean acidificiation" is just another scam. pic.twitter.com/AufWkV57hR ["Steve Goddard" retweeted by Lonny Eachus]

No Lonny, it's not a scam. Extremely ra

https://crypto.stanford.edu/~b...

That's not "my agenda", your 2) meaning is called "faith".

Go and watch Stanford Colloquium where Rick Rashid from Microsoft Research shows a live demo from China translating from English to Chinese in real time. It's near the end (1:09).

You mean near real time...not real time...that would be quite a feat.

Go and watch Stanford Colloquium where Rick Rashid from Microsoft Research shows a live demo from China translating from English to Chinese in real time. It's near the end (1:09).

ML has already been proposed to improve the performance and resource efficiency of large-scale datacenters. Detailed information on two of the most well-known examples from Stanford and Berkeley can be found below: http://engineering.stanford.ed... http://www.eecs.berkeley.edu/P...

And yet, in the next line below carbon fiber is Carbon Toray T1000G at 6370MPa for fibers alone. Then, should you go lower in the chart, you hit 11K+ for carbon variants like nanotubes and graphene. But none of those really matter, because until you place it in a matrix, you won't have anything useful. You'll note the laminate values are much lower. For glass, you don't have that many options better than the standard resin, but for carbon, you can have some interesting materials for matrixes. None of those are listed in the linked chart though. To give you an idea of how much things vary, glance through this randomly chosen report.

On the west coast there's Shenoy's group at Stanford who are doing similar things. Much of their work is currently with monkeys and not sure if they have anything clinically available yet, but at least you can get an idea of what's currently possible:

http://www.stanford.edu/~sheno...

Never mind bothering with calculating the size, as that's not the problem. It's the Latency, Stupid

Also see
http://www.slac.stanford.edu/e...

It was actually done much earlier but generating matter by scattering off of virtual photons. The SLAC experiment was actually looking for (and found) nonlinear interactions in photon / photon collisions. (as were predicted by QED).

Neurons aren't even within several orders of magnitude as fast as transistors: linky1 and linky2.

However, a single typical neuron does a lot more work than a single transistor, computationally speaking.

you should check the studies on this:
http://credo.stanford.edu/docu...

on average, charters are no better. But for some reason, they prove exceedingly successful in educated people in poverty, people where english is not their first language, and special education needs students.

Charters are supposed to be a place where different approaches are tried. And if public schools were doing their duty, they would look at this extraordinary set of results dealing with poor, handicap, or ESL students and find out how they can incorporate this success. There are real successes with at risk populations that are not being effectively serviced by the current public school system. As education is the best way to address the long term issues of generational poverty, I'd say charter schools are proving their worth quite ably.

The other data we have is that charter schools, their first or second year out, are very ineffective. It probably has to do with both the lower experience of the teacher core and trying to figure out what works, or at least that is my random guess based on nothing more than the back of an envelope.

So the best of both worlds would be leveraging the success in dealing with the needs of poor, ESL, and handicap students either as an offshoot or within the framework of the public school system. But no one in TPS wants to hear that they may be failing quite miserably relative to the competition and no one within the charter movement is willing to admit that the data doesn't support the wholesale abandonment of the public school system.

Mesh has it's pros and cons, it can be more secure but the tradeoff is that you suffer a performance penalty the more hops that are incurred. This doesn't give you some of the layer three switching optimizations that are available now in most backbone networks. I'm not saying it couldn't be accomplished but it's not there now. Secondly it doesn't have to be wireless, it could be but it doesn't have to be. I think there's also some great opportunity with the emerging MIMO over Multimode Fiber (MMF) work that's going on which helps solve the traversal of traffic going to a few places. It's in the research stages but probably in the next 5 to 10 years we could probably see an order of magnitude or more in performance over existing links if the research is successful. Of course we'll have to probably wait another 5 years for all the players to come up with standards etc. to properly dominate, er commercialize the tech.

When RFC 791 was introduced, IP was given the ability to enable loose or strict Source Routing. It's just that over the years there's been very few Internet based uses that I can think of where Source Routing is applied in a practical sense and a lot Channel based ISPs (and Network Admins) don't respect it. In fact, I doubt a lot of people realize it's even in there but it gives peers the ability to choose the best route either for security or for performance. It's really up to the peers to decide and for the interceding hops to allow and respect the Source Routing information. The VLAN concept pretty much obviates Source Routing as well since you have a virtual network infrastructure presented may not show any hops at all for example.

Yeah, damned Stanford hippies.

My own version of the preferred policy would be that publicly funded research should be patentable, but the patent royalties must be plowed back into research (not e.g. personal enrichment of the professors).

I think in many cases allowing universities to patent things produces a nice synergy between research and commercialization, which absent patents would result in a lot more use of trade secrets to try to accomplish the same effect. For example, Stanford invented a bunch of early synthesizer hardware; Yamaha licensed it and commercialized it; the royalties from those patents have supported continued research at Stanford's CCRMA. Without patents to help mediate that transaction, you'd have one of: 1) CCRMA tries to build its own synth and get into the hardware business themselves; 2) CCRMA keeps the methods secret and tries to sell them to a synth manufacturer; 3) Yamaha uses them for free, getting a windfall while CCRMA doesn't get anything. I don't think any of those situations would be better.

My problem is with the term of believing. Believing something requires some kind of faith, believing someone requires some kind of trust. Neither has anything to do with science.

Uh...are you sure about that? Just read the first sentence.

Stanford's own Professor Mark Jacobson and twenty Stanford students demonstrated that California could be entirely fossil fuel free by 2050. Stanford should divest from all fossil fuel companies. http://www.stanford.edu/group/...

sweet. Please define free will.

"Free will, even for robots" by John McCarthy:

http://online.wsj.com/news/articles/SB10001424052702303417104579541950544978572

The article written by Bruce Goldman of the Stanford University School of Medicine is a closer source to the original research without being paywalled. It's better than the Wall Street's version; there's less fluff with a little more depth in the explanation and also includes additional links to related sources.

Ineterestingly noted was that this is considered an unsophisticated critical experiment; unsophisticated in that anyone could have done this decades ago without any real knowledge on the workings of the brain itself; critical because of the type results that could be acquired based on the experiment's simplicity in design — it hadn't occurred to anyone to try.