Because what I did in college was write software which among other things was used to model digestive processes. As I recall, the medium sized model was a fairly non-sparse system of more than 200 differential equations, mostly linear but with a few key highly nonlinear terms. And in case it isn't obvious, 200+ equations translates into a hell of a lot more chemicals involved than you could reasonably memorize.

Difficulties of solving the thing aside (the system turns out to stiffer than you'd expect), the fun part for the chemists was getting all the coupling coefficients right, because they come from lots of different sources and are expressed in lots of different units. I built dimensional analysis into the software, but they decided not to use it because it would take to long. That was a bad call, because after publication a mistake turned up (an add that should have been a multiply) that dimensional analysis would have caught. The good news is they reran the model and the mistake didn't change the results significantly, so they were able to issue a simple correction.

I couldn't name even one enzyme involved then or now, but back then I knew quite a bit about how digestion actually worked.

Darn! Forgot to log out and go AC. Sigh.

Writing a post without that symbol in it is difficult, but if you work at it... Voila!


- Gadsby

First of all, your initial assertion that university professors don't get training in teaching is incorrect in many cases. In many places the overlap between the "college of eduction" sorts and the other departments is considerable, with the education professors teaching lots of courses, usually the entry level ones.

And guess what? It's entirely possible for them to be the ones who suck at teaching. That was certainly the case in the department where I taught: Consistently lower student ratings and student performance for that subgroup.

Oh, and as for the non-English speaking graduate students, in a lot of places they are handling entire classes with little or no supervision. Maybe that's still a TA, but when they are the only person teaching the class and doing all the grading, it's pretty much a moot point. And it's a HUGE problem: My calculus classroom used to be SRO because poor guy teaching another section one room over spoke great Russian but almost no English at all. (I say "poor guy" because it is not his fault he was given a job he couldn't possibly do well.)

I'd actually go a bit further, and say that defining slope in the obvious geometric fashion is the better way to do it. It's not like we have a shortage of terms for the change in a car's position or whatever, e.g., "speed" or "velocity". And the more general concept is better referred to as "rate of change".

I used to teach calculus, and even at the college level the grasp students have over the relationship between what that car is doing on the road and those lines on the chalkboard is often fairly tenuous. So a good teacher spends time explaining how the things connect. When you're doing that it helps to have different words for what's on the board and whatever it is in the physical world you're talking about. And it helps if the terms you use connect naturally to the domain where they are being applied.

And as long as I'm nattering on about this, another thing I picked up on pretty quickly is what sort of explanation works best varies from one student to another. For some students geometric arguments work best. For others that's completely opaque, and they get more out of an abstract symbolic approach. There are even some that just naturally get trigonometric arguments. So when I'd do a problem talking about, say a ladder sliding down a wall, I'd typically work at least two ways and sometimes three.

Anyway, it strikes me as more than a little ironic that someone who is supposedly trying to rework curriculum to better suit student needs is so pedantic about terminology. In my experience excessive pedantry is completely at odds with effective teaching.

He's (arguably) transcribing the questions as they are asked and noting his own responses. He is not transcribing anything from a recording. This restriction is clearly intended to prevent a person who prepares a transcript from the recording at a later date from being able to reveal anything. Which is entirely reasonable, although I suspect that in most cases it's more a matter of someone comparing the transcript made at the time with what was on the tape. (At least that's how it worked when I have given testimony. And FWIW, it's been my experience that in relatively relaxed settings like depositions stenographers appreciate it when you spell any unusual technical terms you're using.)

Also look at the subsequent section regarding secrecy, where the person giving the testimony is conspicuous by their absence.

The article in question says, "The most sophisticated and powerful cyberweapon uncovered to date was written in the LUA computer language. Except that it wasn't. The Kaspersky FAQ says:

The effective Lua code part is rather small compared to the overall code. Our estimation of development ‘cost’ in Lua is over 3000 lines of code, which for an average developer should take about a month to create and debug.

Flame is 20Mb total, and a lot of that is almost certainly written in C/C++ (Lua VM, sqlite3, zlib, libbz2).

The article then says, "[Flame] was built with gamer code". Also incorrect. Lua is a general-purpose scripting language in no way specific to gaming. And I've heard nothing that says code directly related to any sort of game is part of it.

That's the last time the article mentions Lua or gaming, but no further mentions are necessary. A false connection has been made, and by hyping up the danger of Flame, e.g., the UN views it is a "significant threat", they're effectively blaming game developers through guilt by association.

The article also says, "Flame came to light when the U.N. International Telecommunications Union (which oversees cyberactivities for the body) received reports of unusual activity." The implication is Flame was responsible for the activity. Again according to Kaspersky, that's not the case. They were attempting to track down something called "Codename Wiper" that was responsible for actually deleting data when they stumbled across Flame by accident.

In contrast, the MSNBC article makes it quite clear that only part of Flame was written in Lua. It then engages in a fairly coherent discussion of why Lua might have been chosen to implement part of Flame, quoting various sources with various different takes on it. And the headline is rather obviously intended to be facetious.

So, on one hand we have a fairly coherent piece that actually tries to get into software design philosophy. And on the other, we have your typical pile of crap from Fox News.

But mostly a lose for patients.

The reason it's a win for doctors is that in our "fee for service" system they don't make much on a simple office visit, especially if they have to take the time to write a prescription. There's also ongoing overhead for prescriptions - whenever you run out of refills the pharmacy has to reauth, which takes up staff time. (This last is why the office where my son works has additional staff on hand on fridays, which is the day that everyone does their refills.) It may also reduce liability, although exactly how that will play out is harder to predict.

It's an obvious win for insurance because most OTC stuff isn't covered. And likewise for drug companies, who will sell more product, a lot of it to people who don't really need it.

The big losers will be patients. Sure, some people will have easier access to the meds they need. But costs will overall be higher, and some of the meds they are talking about aren't all that safe. Statins, for example, may be great at treating high cholesterol, but over time they can cause liver damage. Most of the antihypertensives are reasonably safe when taken at the correct dose, but take a bit too much and things like positional vertigo can occur. And as others have pointed out, the inability to track OTC meds can result in serious drug interactions or overdoses. Indeed, this is already a serious problem with acetaminenophen - so many different products contain it that it's easy to get an overdose, and boom goes your liver.

Irrespective of their quality, Da Vinci's drawings did little at the time to challenge the use of Galen's work (which was based on dissection of animals and therefore quite inaccurate). That particular bit of heavy lifting was done by Andreas Vesalius, who not only debunked Galen, but was also the first to publish a comprehensive work on anatomy (De Humani Corporis Fabrica). His work has repeatedly been found to be highly accurate, especially considering the conditions under which it was produced. An amusing side note is that it was so well regarded it was extensively pirated.

Vesalius made a lot of enemies by going against what amounted to the medical establishment of the time. After repeated challenges his critics actually resorted to the howler that the human body must have changed (evolved? ;) since Galen studied it.

Vesalius has always been a personal hero of mine - a guy who developed an interest in an an important area (anatomy), and pursued it, at great personal cost, with as much thoroughness and rigor as could be had at the time.

This was when we were remodeling. The cost was negligible compared to everything else. It's been about 2/3 full ever since, and I have never regretted it.

The thing I do regret is not running enough cable. I put two CAT-5E in each room and it isn't enough. I should have pulled 4 everywhere. I've had to add a couple of runs, and doing that after the walls are closed up is difficult and expensive.

I've found Wifi to be a poor substitute for wired. When two laptops are backing up trying to watch some video is painful.

Think about it: Cadaveric transplants often come from people who were in an accident. The time available to match the organ in such cases is not going to significantly different than if you grabbed someone off the street intending to steal their organs.

When there's plenty of time, as in the case of altruistic donations, yes, full match testing is performed. But when there's not, it isn't, and it usually works.

In any case, these days not only are transplants done with little or no MHC match (mine was only 2 out of 6), but they are even done against blood type. The process for that is fairly involved, making it impractical to do if you're going to transplant a stolen organ. That said, the odds are over 40% that a person grabbed at random in the US is going to be type O, and type O organs are blood type compatible with everyone.

Actually, end stage renal disease (ESRD) including both dialysis and transplants, is covered by Medicare regardless of age. The only other condition that enjoys this status is amyotrophic lateral schlerosis (ALS). I have no idea why ALS is handled this way, especially since there are several similar motor neurone diseases that aren't covered, but in the case of ESRD, it's because when dialysis was first developed it was extremely expensive and insurance refused to cover it. The result was few dialysis machines were built and the costs remained very high. ESRD was and is very common, so laws were passed to extend medicare to cover it. (And the costs did drop, but it's still expensive.) And when transplants became available, the coverage was extended again to cover that.

There is, however, a little gotcha in all this. The drugs needed to prevent rejection of a transplant are also expensive. And once you have a transplant, you don't have ESRD any more, so your Medicare coverage ends. This was addressed by extending the coverage for 18 months, which I guess is how long transplants used to last. But these days the average is more like 9 years. So what happens is someone gets a transplant that's paid for by Medicare, their meds are paid for for 18 months, then the coverage stops, they can't afford the meds and the transplant stops working. So they go back into ESRD and need dialysis, at which point they're covered again.

This is absolutely insane no matter how you look at it, since the meds are typically around $10K/year whereas dialysis is closer to $50K/year. So not only do you waste a precious organ, it ends up costing more.

The good news is that one of the provision of the AHA that goes into effect in 2014 is to extend medicare to cover the meds indefinitely.

Substantial risk understates the situation if anything. The fact is removing a kidney is a pretty big deal whereas putting one into someone is a lot simpler. This is because they put transplanted kidneys into the lower abdomen inside the muscle layer but outside the peritoneal wall. (The old failed or failing kidneys are only removed if absolutely necessary.) Removing a kidney, OTOH, means going through the abdomen to the other side. Even though it's done laparoscopically, it's still fairly traumatic, to the point where altruistic donors (that's what they are called) have a significantly worse time of it than the recipient in the first couple of weeks post-transplant. Because of this, there is no way in hell any remotely competent surgeon would agree to put back a kidney they are sure she doesn't need so soon after the original procedure. (Donors undergo extensive testing before such procedures. And it's actually surgeons plural, since reattaching blood vessels and hooking up ureters are actually different specialties.)

For that matter, they would not have removed the transplanted kidney from the original recipient were it not for the small matter that according to the article, it was killing him. (When a transplanted kidney fails and another transplant is done they don't remove it unless absolutely necessary, with the result that someone can end up with four or more kidneys.) So they were going to end up with a kidney and no place to put it. Rather than toss it in the garbage, my guess is they started calling people at the top of the list who were type compatible until they found one willing to give it a go.

I'll also point out that one of the side benefits of being a donor is that in the unlikely event that your remaining kidney fails, you automatically go to the top of the transplant list. And in most cases 100% of the donor's costs are paid for.

That may be true for other organs, but not kidneys. And this is for good reason: With, say, a heart or lung or liver, you either get one or you die in fairly short order. But we have an acceptable substitute for a kidney: Dialysis. So, issues of compatibility and availability aside, kidneys are allocated on a first-come-first-served basis,

What this translates to in practice is that if you're blood type O, expect a *long* wait. This is mostly because type Os can only get a type O kidney, meaning about 60% of the kidneys that show up aren't going to be compatible. (This actually isn't true - there's a therapy that allows transplants against type - but since it requires extensive preparation in advance it isn't really practical for cadaveric transplants.) It, OTOH, you're type AB, your wait time is likely to be much less, mostly because any kidney that shows up will work for you.

This approach only works if the messages continue to be sent using the same mix. Which given all the publicity this has gotten and how these sorts of crazies tend to monitor every reference they get in the media, seems very unlikely.

More specifically, they've seized one server, presumably after monitoring it for some time and capturing all the incoming messages. Now they use the private key to re-encrypt the message and look for a match among the incoming traffic. Assuming that traffic wasn't sent using a TLS mechanism with perfect forward secrecy, they now have the IP address of the next to last server in the mix. But what they don't have is any recordings of the traffic getting to that server. And unless the person sending these messages sends some more using the same mix, they will never be able to catch any.

I supppose it's possible that after monitoring the traffic, they also started monitoring traffic coming in to every host that ever sent this system mail. But I'm dubious of the practicality of that, both in the legal and technical sense.

What they should have done is use one of those handy-dandy national security letters or whatever they are called to gain access to the server in secret. They could have pried the private key loose that way, then initiated monitoring on the next server up the chain, another letter, and so on.

Of course this also falls apart if one of the servers is some place that doesn't like the US and won't honor requests from US law enforcement.