In vitro (test tube) tests of lamivudine showed zero efficacy, and the CFR for that doctor's patients is now reportedly up to 5/15, which isn't statistically significantly different from the CFR anywhere else with proper supportive care and no drug treatment. So unfortunately, it appears to have no effect.

However, favipiravir (an anti-influenza drug) shows promise in mouse trials, so it may start to be used in the not-too-distant future. They're also testing brincidofovir. And both of those treatments are readily available, unlike Zmapp. We'll see if either of those turns out to be effective in vivo.

1) Which drug? Do we have enough of it? Can we get it to where it is needed? Has it been previously tested in humans for toxicity? Is there any reason to believe it will work beyond mere hope?

5) Most drugs do not work. Do you REALLY want to spread already scarce resources even thinner on a long shot that probably will not work?

Nothing is being tested without success in animal models. That's not a guarantee, but AFAIK viruses in humans replicate in basically the same way that they do in animals, so odds are reasonably good that they will, assuming that they don't kill the patients or cause other harm, unless the human body filters them out of the bloodstream more efficiently. I wouldn't expect physiological differences to play nearly the same role in treatment for a virus that they would play in (for example) treatments for cancer or Alzheimer's, though I suppose it depends on the method used to disrupt viral replication.

2) We do NOT have particularly good information in the historical record. The medical records in the affected areas are quite certain to be of poor quality. So you lose a LOT of information that is relevant for making comparisons and you do not have a particularly good control group.

That's okay. The records in the test group are likely to be of poor quality, too, so it will balance out. :-D

But seriously, AFAIK, we have a pretty good idea of the CFR of people treated in hospitals/clinics/*. The real record problem is that lots of people avoid going to the hospitals, which means the total death rate may be grossly underestimated, and the total number of cases may also be grossly underestimated.

There's even a minute chance that 90% of people exposed don't get sick, and that the high CFR represents the fatality rate of the worst of the worst. The medical community is relatively certain that this is not the case, but there are no guarantees.

But none of that is really relevant, because they aren't talking about prophylactically treating the entire population; they're talking about treating people who are known to be sick with Ebola, and as previously noted, we have a pretty good idea of what the CFR is among that population.

3) Unless you can control for other variables like public health policies etc you may not know if it worked or how well.

Although that would have an effect on the spread of the disease, it is unlikely to have much of an effect on how likely the infected are to die, short of policies that force people to get treatment sooner (which can be controlled for trivially by comparing people in groups based on how bad their symptoms were when they arrived).

4) Rushing a drug to market can result in losing valuable information about WHY it worked (or didn't) which may be more valuable than IF it worked to future patients

If it is effective in only part of the population, then yes, that's true. But chances are, you can do something resembling meta-analysis after the fact to obtain the same information, as long as your records are good enough, and if they aren't... well, again, they're not likely to be good enough with a slower, more methodical study in those countries, so I'm not sure what difference it makes. :-)

You'd be a lot more convincing if you gave actual reasons that it won't work. But you can't, because you're wrong.

The type of trial that you're claiming has "been tried before and failed" is called an "open clinical trial". They are quite common in the medical field. Although they do sometimes show some bias towards success compared with placebos in situations where success is subjective, when the result of failure is death, there's no real risk of experimenter bias changing the outcome. Either the patient did or did not die. Therefore, your argument is crap (unless your test subjects are cats belonging to Austrian physicists, in which case all bets are off).

Besides, the main reason control groups are commonly used in tests is because most testing happens on a small group of subjects. The smaller the group of subjects, the more critical randomization is at minimizing the impact of small variations between individuals that could otherwise have a greater effect on the outcome than whatever treatment you're testing. When you treat a large enough population of sick people, however, any such variance is likely to get lost in the noise.

In this case, you have a potential test group that makes up a sizable percentage of the total number of people who have gotten Ebola throughout history. Those sick people mostly live under fairly similar conditions, with similar levels of medical care to the people who got it a few months ago. To the extent that this is not the case, you can typically control for those differences when analyzing the data. This makes Ebola a fairly ideal candidate for an open clinical trial.

Now that's not saying that a successful trial means that the treatment actually cured people with absolute certainty; most viruses tend to weaken as they spread, thus resulting in a steadily decreasing case fatality rate over time, so if the improvement is relatively small, there's still a chance of bias caused by differences in the disease over time. With that said, if you suddenly see the CFR drop to 5%, you can be pretty darn certain that the change wasn't caused by a random mutation. So although open trials aren't perfect, if a treatment appears to be highly effective in a large enough open trial, you can probably safely assume that it works, at least in the absence of evidence to the contrary.

Campaign contributions aren't bribes.

The real problem is that what the CU decision effectively made legal were attack ads that mention a candidate by name, paid for by third parties. Although they are effectively in-kind contributions to the campaign, they aren't subject to any of the contribution limits that would normally apply to contributions, because they aren't actually given to the campaign.

In effect, the CU decision allows for unbounded campaign contributions in the form of advertising dollars from corporations, which are prohibited by law from making direct campaign contributions. And unbounded contributions are (effectively) bribes.

If your facts were correct, I'd agree with you, but they aren't:

• This case had nothing to do with the movies themselves. Those are clearly protected speech, and were never in question. This case was about advertisements for the movies aired on TV.
• The FEC did not reject CU's challenge to Fahrenheit 9/11 on first amendment grounds. They rejected it because Fahrenheit 9/11's ads did not mention a candidate by name within 60 days of an election, and thus did not violate the law in question.
• By contrast, they found that CU's advertisements mentioned a candidate by name, and thus were in violation of the law.

CU's challenge was to overturn a law that their ads violated, but Michael Moore's didn't violate. It's as simple as that.

First, I'm not anti-gun. I am, however in favor of limited regulation of firearms to ensure public safety. Second, I'm well aware of everything you just said. However, to play devil's advocate once again, none of those things precludes a fully anti-gun interpretation. Here's why:

Mason's statements indicate what he considered the militia at the time, because the entire people were (at least to some degree) responsible for securing the nascent nation against its enemies. One could reasonably argue that, because the entire people are clearly no longer directly involved in the defense of our country today, his historical definition is no longer relevant.

Similarly, although you are correct that "well-regulated" does mean "properly functioning", as opposed to "controlled by the government", it is the government's responsibility to ensure that such a militia is, in fact, properly functioning; nobody else can do that job for them. Jefferson did not live in an era with automatic weapons, in which every couple of years, somebody snaps and uses them to shoot up a school full of kids. If you interpret "militia" based on that original meaning, then one could reasonably argue that the militia is no longer properly functioning (well-regulated), and that laws must bring it back into line so that it becomes properly functioning again.

In other words, if you want to interpret the second amendment to be absolute, rather than absolute only for people actively involved in service to their country, the onus falls on you to prove that it makes the country, on the whole, safer. Personally, I feel that it probably does, on the whole, except where convicted violent felons and the mentally ill are involved. And I feel that we need better screening to ensure that the mentally ill are not treated as part of that militia, in the interest of public safety.

Ah. They do have an update for that one. Apple didn't release the update through their normal Software Update mechanism because users aren't likely to be affected by it unless they're fairly advanced server users, but they did release a software update for it about two weeks ago. You can download it here.

OSX Yosemite is still waiting for a patch for Heartbleed...

Before they can ship a patch, Apple would first have to ship an affected version of OpenSSL. For binary compatibility reasons, Apple never moved its OpenSSL libraries off of the 0.9.x branch. Currently, Yosemite ships with version 0.9.8za. The first version of OpenSSL affected by Heartbleed was 1.0.1.

...which means I have to keep a large number of services I'd like to be running disabled for the time being.

In other words, unless you installed an affected version of OpenSSL (in which case you will have to install an update yourself), it is safe to turn those services back on, and it always has been.

Ebola would be a terrible bioweapon for the simple reason that it isn't airborne or waterborne.

Bear in mind that the CDC uses a very narrow definition of airborne that effectively limits the term to viruses that cause a lot of sneezing. With that said, any time you have a viral load of any kind in the sinuses or salivary glands, that virus can theoretically be spread through the air, and Ebola is no exception. It isn't very easily spread through the air by humans because humans don't get much of a viral load in their noses and salivary glands until they are extremely sick. With that said, that doesn't mean you can't get it through the air; it just isn't particularly likely unless you're sharing a confined space with someone who is infected.

Citizens United was not the correct ruling. Bribes are not a protected form of speech, and large campaign contributions made by an individual or a small group of individuals (remember that in a corporation, a small number of board members have near total autonomy in such decisions) are indistinguishable from bribes. And even if bribes were a constitutionally protected form of speech, it would still be entirely reasonable for government to limit the manner of that speech—requiring corporations to spread out those contributions over a longer period of time, to level the playing field. Those rules were quite clearly a "reasonable time, place and manner restriction". And I say that even as someone who is a staunch defender of the near-absoluteness of the first amendment.

The entire decision on Citizens United was based on absurd hypothetical scenarios (e.g. the government twisting election laws to censor the media) that did not apply to the actual case at hand in the slightest. That decision was a clear case of legislating from the bench, in the worst sense of the phrase.

So no, SCOTUS doesn't almost always make the correct ruling. Its shortsightedness frequently results in fairly serious errors (Plessy v. Ferguson, for example). Fortunately, it usually corrects those mistakes eventually. In the same way, Citizens United will likely eventually get overturned, assuming the piles of corporate money don't hopelessly corrupt the government beyond redemption before such a correction has a chance to happen.

I agree with you on the NSLs, assuming they aren't time-limited. If they are time-limited (e.g. a five-year embargo), it becomes a much harder argument; you'd have to argue that the time period was not reasonable, thus resulting in an unreasonable prior restraint of speech.

As long as the question "How would we know" is illegal for the participants to answer, we should simply assume the worst.

This. That's why the first amendment is by far the most important aspect of American law, and that's why any laws that abridge it in a broad manner (as NSL gag orders do), no matter the perceived need, must be struck down. There can be no freedom without the true transparency enshrined in our highest law, and without the people having the courage to demand that such transparency be defended above all else.

Exactly. The second amendment is arguable. Depending on how you interpret it, it may or may not have been intended to apply to random people having weapons for their own personal protection. Therefore, it is at least possible to argue that the mere existence of police departments, national guard units, and the military are well-regulated militias, and that nobody outside of those organizations has a right to bear arms. I'm not sure I necessarily agree with that interpretation, but it is at least possible for a rational person to argue that interpretation of the second amendment without looking insane or idiotic.

By contrast, there's no way to interpret the first amendment in a way that gives the government any right to pass laws that cause prior restraint of speech of any kind. With that said, courts have generally ruled that certain limitations on speech (as opposed to outright bans) may be allowed if they qualify as "reasonable time, place, and manner restrictions." For example, the government can legitimately say that no protests may occur within two blocks of a cemetery, that no stores can sell alcohol within three blocks of any elementary school, or that no protests can use a megaphone between 10:00 PM and 7:00 AM.

A national security letter's prohibition on disclosure, so long as it is time-bounded, might legitimately fall into that category. If those gag orders have such a time bound, the EFF may have a very difficult time arguing the case (unless the time bound is long enough to be seen by most people as unreasonable). With that said, I'm not under the impression that NSL gag orders have any set date on which the embargo is automatically lifted. If that is the case, then such prior restraint is clearly and unequivocally an egregious violation of the first amendment that cannot possibly be justified under any remotely plausible interpretation of the Constitution or its amendments.

Of course, the fundamental problem is that most people taking an oath to uphold the Constitution against all threats, both foreign and domestic, automatically assume that this means protecting the United States against its enemies, never even considering the possibility that the actions they take in doing so represent an even bigger threat. Throughout history, nearly all of the worst atrocities have been under the guise of protecting the people from its enemies, from the Crusades to the Holocaust. Worse, most such horrors often start out in a manner that seems benign—Hitler and McCarthy, for example—and get progressively worse. National security letters might start out benign, with the best of intentions, but they have the potential for nearly unlimited amounts of abuse, which is why IMO they must be found unconstitutional sooner rather than later.

To that, I would add one more point—that the first amendment is by far the single most important amendment to the Constitution, without which the rest of the amendments are moot—the condicio sine qua non, if you will. It represents the very core of our freedoms. Without the ability to talk about and expose abuses, there is no way to prevent them. Without the ability to criticize the government, there is no way to change it. Without knowledge of what government is doing, the entire notion of democracy—rule by the people—becomes untenable. It is a near certainty that any such government will rapidly devolve into a thinly veiled plutocracy or aristocracy, because the people cannot adequately evaluate their lawmakers in the absence of any understanding of what those lawmakers are doing. For this reason, the first amendment must be defended at any cost, above all other amendments. And that is what the ninth circuit must bear in mind as they make their decision on this matter.

Or just run each app in its own VM so that when it turns rogue, you can cleanly shoot it in the head without any widespread damage.

Actually, I do know how close to reality those numbers are. They're not close at all. The odds of an fatal accident are less than 1.3 per 100 million miles traveled, and the odds of any accident are about 231 per 100 million miles. So you're off by more than a factor of 4 in your accident rate per mile. And even though there's a lot of correlative "evidence" that suggests cell phones cause a lot of accidents, there's much stronger evidence suggesting that this is not, in fact, the case. In fact, the numbers actually suggest that the opposite may be true—that increased cell phone use has reduced the rate of accidents.

Over the past couple of decades, as cell phone usage grew from essentially nobody having one to roughly everyone having one, the number of accidents per mile has been steadily decreasing. This suggests that in the grand scheme of things, either cell phones have no appreciable effect on accident rates, or that any effect that they have is more than negated by other factors, ranging from better braking and traction control to the extra cognitive ability resulting from people doing more multitasking in their daily lives.

Now this is decidedly not the case for teenage drivers. They still suck at driving, and as a result, it is quite possible that the extra risk from reading that text will be greater than the reduced risk from the extra miles. We shouldn't allow teens to use cell phones while driving for the same reason that we don't allow them to drive with other kids in the car—every little distraction is a problem if you lack the driving experience to do it instinctually. But the numbers strongly suggest that such prohibitions are completely ineffectual when applied to the population as a whole. (Although that's about the California law, similar studies have been done in other states that have passed anti-phone laws, and AFAIK, they've all consistently shown that such laws have no statistically significant effect on accident or fatality rates. None.)

In other words, the numbers agree with me and disagree with you. The cell phone distraction myth is just that: a myth. Glancing at a cell phone and reading a short message is no more dangerous than glancing down at the radio or the air conditioner or any of the other random things in your car that you might look at for two or three seconds. As risk factors go, it is almost completely lost in the noise.

What you're fundamentally missing is that the increased risk associated with skimming a text is over a very short period of time. The total risk requires you to multiply that risk times only the portion of your trip that it affects. Even if reading that text made you 100x more likely to have a wreck during those three seconds out of a twenty minute trip, the total risk averaged over the trip as a whole would still be somewhere around a 25% total increase. By contrast, if not taking that call means that you drive 30% more miles, the odds of you being of an accident increase by 30%, because accident rates are roughly linear in the number of miles driven. And I'm not convinced for one minute that it makes you 100x more likely to have a wreck, because for the overwhelming majority of those 3-second periods, there will be nothing happening in front of you whatsoever. I'm not convinced that you're even 10x more likely.

But suppose we go with your numbers and assume that your odds of an accident per mile is 1 in 100,000, and skimming a text gives you a whopping 1 in 10,000 chance of a wreck. That means that if reading the text saves at least ten miles of driving, you're still better off reading the text than not reading it. That's not a particularly high bar. The average American has a 25.5 minute commute each way, so assuming you're equally likely to be asked to stop at any point along that route, using your numbers, on average, you're still better off reading the text message than not reading it, assuming you get it near the beginning of your trip.

No Belkin, no LinkSys... who's still making routers that you can find in regular stores?

D-Link. Anecdotally, they seem to be some of the most solid gear I've used. Every time I've tried anybody else's hardware, I've regretted it.

For example, I once bought a TRENDnet green switch. It probably would have worked just fine under ideal conditions, but it failed to fall back to slower-than-gigabit speeds correctly through a cable with a bad pair (even when I told my upstream Cisco managed switch to advertise 100-megabit or 10-megabit speeds). I returned it and bought a D-Link, and because it actually connected (at a slower speed), I was able to instantly diagnose the problem, whip out a cable tester, track the problem to the wire in the wall, crack open the wall panel, redo the punchdowns, and fix the problem.

For another example, I once bought one of those white Netgear 10/100 switches. A few minutes after I plugged it in, I noticed that I was having network connectivity problems. Upon further testing, I discovered that I was seeing double-digit percent packet loss across that switch. I yanked it and replaced it with a 10-megabit hub until I could return it and swap it for a D-Link, which worked flawlessly. (Incidentally, not long thereafter, I encountered that same model of Netgear switch at work, with one of our printers attached. Every so often, it just stopped passing traffic until you unplugged it. In other words, not a fluke.)

At this point, if it doesn't say either Cisco or D-Link, it isn't going into my house.