Clean room reverse engineering and reimplementation.
One team of tainted developers reverse-engineer the code, and write a spec. Another team of *untainted* developers only get to see the spec, and write code that conforms to this spec.
At least that's how I understood it. Basically, they have a legal way of reimplementing things, including any hidden APIs. Of course, in the US, the DMCA might make the first part illegal, but that part can be done elsewhere if that is the case, and I don't think the spec is legally tainted by any infringement of the DMCA, at least not if it's done outside US juris(male)diction.
Why not just go with the approach of giving a daily fine, like the EU wants, until they open up their stuff?
And, to make there be some *real* incentive their, force them to make these fines payable directly to a foundation that essentially acts to fund ReactOS and WINE development?
That'd be a bitter pill to swallow, and a win-win situation for everyone else.
This may be redundant by now, but the issue is really simple.
This is an alpha release. As such, they're still in the midst of the code-test-debug-lather-rinse-repeat part of the cycle. That means kernel panics. Lockups. Debugging and single-stepping. The works.
Hence, almost all the developers will be running it primarily in a virtual environment, so that the work can progress at a decent pace. This makes sense.
Also, it's not yet a replacement for Windows, which means there is no real reason to focus on those aspects as of yet, so people spend their time trying to get it to a point where it starts to be a real alternative, and when they reach that point, they'll start concentrating on real hardware.
It just wouldn't make sense to spend their time trying to get all the details of a real machine working right now. They'd get bogged down with the boring stuff, not producing anything that is of real value to anyone, and not making visible (and motivating) progress towards the headlines.
Once you've fully replaced windows on a virtualized platform, getting it to work properly on a real platform is a lot less hassle, and a lot more motivating. And motivation is a real factor in any open source work; more-so when it's something as specific and formal as dealing with a clean room reimplementation/ripoff of one of the biggest milking cows in the industry.
If they work on that "real hardware" part now, they'll be doing lots of gruntwork to get something not-quite-functional running on a poor development environment, meaning no real payback and no real motivation. If they work on getting the "not-quite-" bit out of the equation, they're seeing real payback on their invested time all the way, which motivates contributors to get stuff done.
Once it's a functional solution, but limited to virtualized hardware, getting it onto real hardware suddenly *has* a real payback, and the motivation factor is back into the picture, with the added bonus in the form of a "zomg we're almost done" headrush.:P
Agreed. Now all it needs is built-in XMPP support and the Gtalk extensions to that, plus an orthodox file manager and an nvi-clone, and I'll be all set to spend most my time in Opera.:P
OT, I'm not from the UK at all, nor from the US, though I have my fair share of gripes with my gov't.
A power leak is intrinsic to government, or indeed just about any process, I think. As technically savvy people, we know this already. Processing expands to fill available CPU power, memory and disk in a non-linear relation to the amount of work done. Management grows entirely independently of the amount of stuff to be managed.
Also, there is the factor of entropy, which also appears to be intrinsic to higher-level processes, not just lower-level ones. Just compare slashdot now to, say, five years ago.:P
The problem is not so much getting a large group of people directing their guns at something. It's getting them to agree where the guns need to be pointed, and what needs to be done once the dust settles. You could go out and get yourselves a bunch of gun-toting revolutionaries today, and start killing off government people today, for all I care.
The bottom line, however, remains: there's a *seriously* large support machine at work here. You'll be seen as displacing a legitimate government in an insurgency, so you need to either retain power by force, from which it will be really hard to step down without further bloodshed and possibly retaliation, or you need to turn public opinion, which pretty much entails turning your guns on the media and other support infrastructure. That's some serious carnage, either way.
If you don't commit to that carnage, stuff will bounce back even harder, with people crying out for even more stringent measures than before, to protect their illusion of safety and "normalcy". And if you do, there's at least decades of work to be done in order to stabilize the country again.
Also, there's the whole issue of whether all these gun-toting revolutionaries actually agree on things afterwards. You'll quickly see that you disagree on many points of how things should be run. In the start, when you're working on the simple stuff, everything will go smoothly (for some very stretched definition of smoothly), but once you're getting down to the details of it, there will be a lot of cooks in a kitchen that's getting smaller by the minute. It takes a real strong leader to hold something like that together, and then you're really no better off than before, since it'll be that leader's vision, or it'll all fall apart.
As for the constitution, the real problem is that the population isn't cut out for it. Now, I'm not saying this is a USA-specific problem, far from it. But the constitution postulates an aware populace, which is willing and able to think things through in a Big Picture(tm) way and then follow through on a local level. If you had such a population, things wouldn't be where they are.
Regarding the colony stuff, that's some very different circumstances. Try the civil war or something like that for a better comparison. Or the revolution against the Romanov's. And even in those cases, I suspect the resistance was far less than you'd encounter with a revolution now.
People cherish their lives as they are. They're not all that concerned about liberty or anything as vague as that. They have enough of it that they don't feel it is a pressing issue. Sure, a few will complain here and there, but when it comes down to it, they're more worried about their X-mas bonus and who wins American Idol or whatever.
I mean, come on, how many people here on slashdot have voiced the opinion that someone should take up arms and Change Things Now(tm)? Where's the mailing lists, or what-have-you, organizing this revolution? If you really feel that strongly about it, why aren't you preparing a revolution already? Are you armed? Ready to start shooting fellow citizens? At the very least, you could take a shot at the president or something, send a message that the time to act is now or whatnot.
It's the old frog on the boiler plate again. You're seeing that things aren't what they should be, but that's been the case for a very long
You seem to forget... guns aren't about changing governments, they're about removing them.
Sure, you can remove a government with enough force-of-arms, but how are you going to go about setting up a new government that is better than the old one?
This is a government that represents a significant percentage of the population (26%, I think someone here said). That means 26% of the population is opposed to a revolution, and you'll have to supress them or kill them. Not generally considered a good way of governing a country, now, is it?
This *beautifully perverse* aspect of democracy is too rarely appreciated: in a revolution, there is no longer a single target. A revolution must target the masses. And violence targetted at the masses is something you'll need some pretty hard and pressing reason to make people resolve to do on their own.
Quite simply put: in a democratic society, you can't have revolutions unless *any* alternative is better than the current situation, and for the majority of the population at that.
Social hygiene or potential future catastrophe just won't cut it. It has to be a problem that has *already* materialized, and is significant enough that people are willing to risk their loved ones waging a bloody civil war over it.
There is no established link between Al Quaeda and Iraq. And people pulling out is mostly a matter of not wanting to be in the middle of the cesspool aftermath of the US invasion. You know, civil war and all of that, not to mention the country falling into the hands of religious extremists and being more likely to ally with Iran etc.
And that Northern Ireland is still a part of the UK is a testament to the effectiveness of seeking other solutions than bombing someone you don't like.
Also, as Chomsky has pointed out, terrorism is not effective for the weak. It takes something on the order of an actual nation with the attendant resources to make terrorism work. It's worked great for the USA over the years. Not so for smaller groups.
Again paraphrasing Chomsky, the only thing that was special about 20010911, apart from the media coverage, was that (for once), terrorism worked out well for the weaker party, kind of like David and Goliath. And, of course, that the US response triggered a significant growth in terrorism activity world-wide and gathered more support for the terrorists, although I guess you could file that under "worked out well".
With regards to people taking antidepressants because their lives aren't perfect, I always found that rather funny.
Regular antidepressants (the ones you'll get coverage for in an outpatient setting) don't actually have mood-brightening properties, and many are a bit passivizing. Virtually all of them have some pretty nasty side-effects. So they're perfectly useless for people whose lives just aren't perfect. It makes sense that some people get prescribed antidepressants by ignorant doctors without knowing that they'd be better off with a piece of candy, but I haven't seen any hard numbers on this.
As for major depression, where these drugs have an indirect mood-brightening effect by reducing the depressive symptoms (just as a cure for cancer would have definite mood-brightening effects on the people who have cancer), these doesn't necessarily have anything to do with your life not being perfect. And they do entail your life being, in slashspeak, "teh suck". More so than anyone who hasn't had it, or lived every day for a long time with someone who does, are able to comprehend.
That's kind of been my guide to evaluating whether someone should be on one or not: if the adverse reactions and/or side-effects are something they can't handle, even if the drug has an effect, then they shouldn't be using antidepressants in the first place. It just isn't serious enough to merit treatment, and they need to build coping skills instead, or see a lifestyle guide, or possibly go to therapy (or their priest or whatnot, which does as much good as regular therapy, studies say).
Personally, I've had side-effects that a modern person would consider debilitating by themselves, as well as adverse reactions that were grounds for hospitalization (e.g. circulatory collapse). I still don't see it as a problem: the baseline is worse than even the worst adverse reactions. Compare with cancer patients who get to puke their guts out, lose their hair and suffer lots of pain... if there is any chance of getting better, they generally don't cut the chemo or whatnot.
Killing a superhero with a sniper-rifle sort of makes a statement. It isn't fitting. For anyone.
The realities of our time caught up with him. Cap is dead because the concept for which he was a symbol is dead. And what better way to drive the point home than to be brought low with a sniper rifle?
As for the site you quoted, they seem to be missing a few vital points.
For instance, they are criticizing the technology as being experimental. No-one is arguing against that. Call him an "early adopter". Field use in clinical practice is the only way to get research done that is targeted at clinical use, rather than at maximizing profits. Plus, showing distinct differences in organ function can help get people to accept the legitimacy of illness and such. Showing people the reason why compliance is important can also be a good thing.
The language is criticized, blatantly ignoring that the language they are critizising is aimed at patients, not professional. Wordings like "brain balance" are a lot easier for a regular patient to relate to than intricate discussions of the topography of the brain and so forth. Similarly, saying "soul" better describes to the patient what you're referring to than any intricate philosophical treatment of the topic of what comprises the essential features of personality and cognition before launching into the various theories on which features correlate with what in the brain.
I never tell the PHBs what kind of algorithms, schemas, etc. I'm using, unless they ask. Because that's my job, not theirs, and I should explain things to them in terms they can understand, not blather on in techspeak. If they show an awareness of the subject, and an interest in further details, I'd be perfectly happy to level with them, but no way am I going to market myself on technobabble.
For efficacy comparisons, that's pretty simple. If anyone out there that doesn't have a vested interest, but does have access to SPECT, wants to, they should have no problem doing a serious trial. Of course, the fact that e.g. the drug companies don't do this, could indicate that a lot of people are quite uncomfortable with the idea of potentially diminishing their own sphere of influence by indicating that their popular theory isn't all-encompassing.
As for patients doing better, that could be evaluated. The ones I know of have had little or no success elsewhere, but that's just anecdotal again. The identified depression subtypes show a lot better correlation with what I've observed (trait clusters, med response) than any other prevailing theories that I've seen.
False positives and negatives: the critics are probably in a better position to test this. Amen Clinics don't get enough patients fitting regular inclusion/exclusion criteria to evaluate this themselves.
Whether these functional changes are primary or secondary is also most likely irrelevant. Abnormal perfusion and/or metabolism indicates a potential treatment target. Whether the abnormality is due to the illness or causing it doesn't really matter, as long as the treatment works. For research, it matters, of course. A lot. For treatment, the results matter. But, yeah, he should post some more data about any clues he might have about this. My guess is the astroglia are involved in long term downregulation of perfusion and/or metabolism as a result of abnormal activity in those areas, accounting for the slow normalization over time. But some data would always be nice. It would be nice to have the same concerns addressed for MRI, or any of the popular research techniques, too.
But, yeah, David Amen has taken it all a bit far, and has an obvious, financially motivated agenda. So does virtually everyone else. I think the tests are useful in relation to depression. And I have no ties to the clinic, nor have I ever been treated there, nor received a SPECT elsewhere. YMMV.
There is some small core of truth to his statement, although not the way he put it.
Psychotherapy, with the exception of Cognitive Behavioural Therapy, is generally not particularly effective in dealing with "true" depression (which excludes stuff like bereavement, etc), and a "good listener" will do as well as anything but CBT. This has been verified via extensive testing.
MRI is not a very effective tool, compared to PET and SPECT.
Inclusion/exclusion criteria for drug trials leave you with patients you will virtually never see in clinical practice, making the efficacy numbers useless. This also causes issues with women, african-americans, etc.
Drug trials almost exclusively rely on drug-placebo or drug-tricyclic comparisons. It is rare to see a study that is statistically useful, i.e. has drug, placebo, comparable other drug, drug of different class.
SSRIs have no statistically significant difference in efficacy, time to onset, or any other parameter of relevance, aside from CYP enzyme profile. Their efficacy is very limited.
SNRIs have no statistically significant advantages over SSRIs, but more severe withdrawal symptoms, although there is some weak evidence of shorter time to onset.
TCAs are not very viable, due to intolerable side-effects and cardiotoxicity. Their efficacy is variable.
DRIs work very well on several subtypes of depression. Statistically significant advantage in most cases. Fast onset. Abuse potential. Not available, due to the latter point.
MAOIs work very well on most types of depression, if the right type is chosen. One study shows 85% success rate in treating refractory depression with tranylcypromine at doses of 1.3-2.4mg/kg/dy or so. Onset varies from hours to weeks, depending on the type of depression and the choice of MAOIs. However, they require an intelligent doctor and an intelligent patient to be safe, so they are not viable for mass marketing due to the shortage of both.
etc...
I could go on, or even cite this stuff, but this is/...
Bottom lines:
- Psychologists are often full of crap, and charge exorbitants of money for something they themselves have proven equivalent to a free conversation with your minister or what-have-you. Not counting the CBT people here, of course; there's ample evidence of efficacy for CBT.
- Doctors, and even psychiatrists, are commonly not really up-to-date in their own field, and lack familiarity with any good tools other than what in programming terms would be called shotgun debugging. Doing research in clinical practice is frowned upon, despite being the only way to get real-world applicable results. Using "experimental" tools, such as SPECT, is frowned upon, despite a proven track-record.
- Drug companies develop drugs for profit and damage control, not efficacy. That is, they are designed to do the least amount of short-term harm in the hands of J. Doe for the least amount of production expense. Never mind osteoporosis, asthma issues, weight gain, etc, as those fall outside the scope of the FDA approval process, being long-term and all.
- Simple base research is omitted, particularly where it might threaten the profitability of meds, even competitors' meds.
- Depression sucks, and will continue being the featured syphilitic blow-job of our times for the forseeable future. (WHO projections are rather pessimistic. Mortality rates aren't exactly great, and don't seem to improve. And, like syphilis, a lot can be done to prevent/treat, but isn't.)
That's SPECT, not MRI, but it shows this stuff pretty clearly. Compare with some of the other stuff there if you like.
You will not see structural changes. You will, however, see differences in metabolic activity, etc.
And, yes, these can be used as a diagnostic tool, as well as an aid in selecting proper medication to deal with the problem, which will often be necessary by the time you get such a scan.
The condition is not defined by subjective feelings. Have a look at some SPECT scans, for instance.
Subjective feelings are frequently employed in making a diagnosis, however.
There are specific alterations in neurotransmission patterns, glucose metabolism, and other objective, measurable features of central nervous system function. And there are measurable psychomotor and cognitive deficits that develop over time.
Also, the animal models generally reflect a set of fairly objective behavioural features that we can use in working with depression. Essentially, it is most readily induced in animal studies by making a situation hopeless, negative and unavoidable over an extended period of time. You give up, and lie down to die, in a sense. Nothing matters, or is worth doing, because you're just waiting to die anyway, and there's nothing you can do about your situation at your current energy level (or, by the time there is, you have lost too much function to be able to rise to the challenge without extensive rehabilitation).
These are the animal-model features, and their presence can generally be verified in a fairly objective manner. There are also scales that quantify these features in a manner that shows a significantly positive correlation with prognostic markers, etc...
Interestingly enough, many commonly employed scales have very poor severity resolution ability in the to end of the spectrum. When I maxed out a number of more common scales, the MADRS scale had the ability to track variations with a good correlation to the subjective feelings of severity as ranked by people living with me and objective impairment in function and cognition. At all scores from about 20 through 50 on MADRS, I have usually retained a constant near-max score on HAM-D, except my HAM-D improves if I'm too screwed up to get out of bed, or even move, over an extended period of time. Of course, those are the times when being dead would be an objective improvement if only considering oneself.
I think an FDA-mandated scale with a solid mathematical basis (factor analysis, normalization, etc), and good resolution throughout the spectrum from having-a-great-day-every-day to stark-raving-mad-and-needs-weeks-in-the-ICU-before -treatment-begins, might have a pretty interesting impact on future developments in the field, particularly with regards to clinical practice and drug efficacy trials.
Have a look at the Amen Clinic pages. They show SPECT images of patients with various mental problems, including depression. Also, they identify several characteristic types of depression, based on the SPECT findings, that respond to different classes of drugs.
The most visible depressions to someone "on the outside", are light depressions (when the mood features are dominant, and there is still energy to bitch about it) and severe depressions with psychotic features (when the whole lost-touch-with-reality, killing-yourself-for-spilling-some-milk, no-rationality-left kind of behaviour takes over).
To me, I've generally been crying more in my better periods, when I have feelings, and actually have any tears left to cry. In a bad period, I can cut myself deeply and not even feel any physical pain, let alone emotional response. If my sister had died during her recent hospital stay, I wouldn't have felt anything. And I've always cared deeply for my family. I would grieve during the subsequent better period, though.
Crying is a good sign that it's not "too severe", unless it is accompanied by psychosis. By which I don't mean to say that spending your days crying in fetal position is a good thing, just that it's about halfway along the spectrum of severity in depression. Which should give people some idea of how bad the really bad cases are, or how comparatively "light" the more "common" depressions are (and how this leads to statistics about full recovery in a few months, etc.)...
Depression causes a general, progressive decline in cognitive and (later) psychomotor function.
This decline slowly starts to reverse after remission has been achieved.
My therapist did research on this, and the results are pretty conclusive.
Trying to link depression to a hippocampal function phenotype might go places, but I seriously doubt it. And their study appears to essentially not be properly equipped to differentiate general decline in CNS function from anything related to the putative phenotypes. As it is, they're just measuring a decline in hippocampal function.
However, the study can be useful, if taken the right way. Basically, it may allow the development of another objective metric for cognitive decline, which may be useful in diagnosis and determining the severity of the animal-model features of depression (as opposed to the mood related features, etc).
Now, if they could just come up with a better prognostic tool for determining the class of antidepressant that will work for a given patient... Preferrably without needing to resort to a SPECT analysis or somesuch.
Come to think of it, there are some pretty good prognostic markers, but the real problem is that it takes ages before state of the art becomes common knowledge in the research community, and ages more before this becomes common knowledge among psychiatrists, and forever before it becomes common knowledge with the general practicioners.
That, and a lot of patients cannot be treated in the current political environment because they are dopamine-deficient. And the opioid-deficient ones are even worse off, quite often.
Because, that would defeat the purpose of having a handy DDoS-from-hell service around where anyone can anonymously hit tech targets they dislike with 1 day downtime.
This is not a matter of whether they got their monopoly through fair dealing, or even whether the competition is good enough.
It's a matter of making competition *possible*.
As long as they have an effective monopoly, they are not allowed to lock others out of the market. They do not have to actively *help* their competitors, but they may not block them from competing.
Let's take an example. Unless an office suite is virtually 100% compatible with Microsoft Office, it cannot fly, in the sense that it can never achieve significant market penetration. If you build a suite that is 100x as "good" as MS Office, but lacks interoperability with the legacy of the MS monopoly, it will not succeed.
Now, by preventing you from interoperating with MS products, MS is preventing you from competing. Not by doing a better job than you (winning the competition), but by leveraging their position as a monopoly (blocking market entry). The latter is detrimental to the market, and illegal in some areas.
By forcing them to allow competition (by e.g. demanding that interoperability be possible for competitors to achieve via availability of specs), you are not helping the competitors compete, you are making it possible for competition to exist. The potential competitors still have to pull off the competition part of it themselves, building the software, marketing it, etc..
The difference is that with forced interoperability it is *possible* for a competitor to succeed, if they do a good job, whereas without it, it is *impossible* for a competitor to succeed.
Grab the sequenced junk DNA, do some entropy analysis etc to compare it to the functional genome. Anything that turns up highly unusual levels of entropy (indicating compression) would be suspect.
There may be some truth to this, but there is nothing near that level of coordination in the IT world at present. If there were, I'm pretty sure something would be done about it. Fast.
Either way, giving any indication that you are aware of and/or have considered this line of thought, is a pretty surefire way to make them want to tighten your leash, or fire you outright.
It was designed to do this. It was implemented. It worked. Just how any mission-critical software should be.
Any production grade high-availability system will do this. And designing a system to be HA is not as close to brain surgery as a lot of PHBs and codemonkey cowboys make it look.
The problem is with the "moral core" concept. Basically, humans tend to seperate beings into two groups, with one being "protected" by the moral code, and the other not being protected. Shifting the boundaries between these two groups is an important part of getting soldiers to fight, and the critical component in allowing genocide to happen.
Most people tend to put animals outside their moral core. Familiarity and empathy extends some limited protection and benefit to household pets in most cases, but even these are not usually part of the moral core.
The scary thing isn't that they are doing the same *things* as the Nazis, but rather that they are doing so based on the same *line of thought*.
Off-topic, I wonder how many intelligent sadists fly under the radar by seeking out positions such as these, where they can cook up some flimsy excuse to perform legally sanctioned torture for very little real benefit. I mean, sure, the pigeons can probably be used for something (e.g. delivering a small payload, or performing inconspicous reconaissance), but the technology to do these things, including inconspicous reconaissance, is probably available in China without resorting to this.
... isn't that a corporation ignored potential national security concerns.
What alarms me, is that apparently Lockheed-Martin did.
I mean, if the single SCMM level 5 company could get their shit stolen, then either something is seriously fishy in them not having infosec to match their coding practices, or someone actually told them security isn't an important parameter for this assignment.
What makes you think drug addiction should qualify for this kind of thing?
For one, the dopamine-response associated with CounterStrike is more than an order of magnitude greater than what you get from e.g. heroin.
Second, opening the door to dehumanizing and mistreating *one* group of people means accepting the mental mechanism (moral core) that allows stuff like the holocaust to happen. Now, this isn't meant as a reductio ad hitlerum argument; I'm just pointing out that it's the same mental mechanism that allows the slope to get slippery, and that it's been shown (repeatedly) to be trivial to alter the perceptions that define this border (inside/outside moral core).
Not that I'm doubting that a lot of kids get labelled with ADD, ODD or what-have-you without fitting the criterion for this, but your argument makes no sense.
Giving ADD-medication to a regular kid that's all over the place will make that kid *more* impossible to manage. We're talking about fairly strong CNS stimulants here. They increase psychomotor activity in "healthy" kids. A lot.
Perhaps what you meant to say was that we have an unfortunate tendency to give ADD meds to the kids when they don't actually need it? I can go with that. The problem here is informed consent: they can't give any without having tried it in advance, and that requires having the parents make the informed consent for them, and docs don't always remember to pull the kids back in to ask them if they prefer life *with* the meds or without.
Oh, and, personally, I'd worry more about how Risperdal has been approved to treat irritability, agression and other signs of not being happy, content or "socially well-adjusted" among kids with autism spectrum disorders. That's an antipsychotic, and it might take years off their healthy adult life: statistically 25 years if you take it throughout your entire life. If that effect is one that accumulates linearly, that means the kids lose 1 year of life per 2 years on the drug. I'm not at all convinced this is a tradeoff they'll be happy about later in life.
Slashdot Mirror
I think you're missing an important point:
Clean room reverse engineering and reimplementation.
One team of tainted developers reverse-engineer the code, and write a spec.
Another team of *untainted* developers only get to see the spec, and write code that conforms to this spec.
At least that's how I understood it. Basically, they have a legal way of reimplementing things, including any hidden APIs. Of course, in the US, the DMCA might make the first part illegal, but that part can be done elsewhere if that is the case, and I don't think the spec is legally tainted by any infringement of the DMCA, at least not if it's done outside US juris(male)diction.
Why not just go with the approach of giving a daily fine, like the EU wants, until they open up their stuff?
And, to make there be some *real* incentive their, force them to make these fines payable directly to a foundation that essentially acts to fund ReactOS and WINE development?
That'd be a bitter pill to swallow, and a win-win situation for everyone else.
This may be redundant by now, but the issue is really simple.
:P
This is an alpha release. As such, they're still in the midst of the code-test-debug-lather-rinse-repeat part of the cycle. That means kernel panics. Lockups. Debugging and single-stepping. The works.
Hence, almost all the developers will be running it primarily in a virtual environment, so that the work can progress at a decent pace. This makes sense.
Also, it's not yet a replacement for Windows, which means there is no real reason to focus on those aspects as of yet, so people spend their time trying to get it to a point where it starts to be a real alternative, and when they reach that point, they'll start concentrating on real hardware.
It just wouldn't make sense to spend their time trying to get all the details of a real machine working right now. They'd get bogged down with the boring stuff, not producing anything that is of real value to anyone, and not making visible (and motivating) progress towards the headlines.
Once you've fully replaced windows on a virtualized platform, getting it to work properly on a real platform is a lot less hassle, and a lot more motivating. And motivation is a real factor in any open source work; more-so when it's something as specific and formal as dealing with a clean room reimplementation/ripoff of one of the biggest milking cows in the industry.
If they work on that "real hardware" part now, they'll be doing lots of gruntwork to get something not-quite-functional running on a poor development environment, meaning no real payback and no real motivation. If they work on getting the "not-quite-" bit out of the equation, they're seeing real payback on their invested time all the way, which motivates contributors to get stuff done.
Once it's a functional solution, but limited to virtualized hardware, getting it onto real hardware suddenly *has* a real payback, and the motivation factor is back into the picture, with the added bonus in the form of a "zomg we're almost done" headrush.
Agreed. Now all it needs is built-in XMPP support and the Gtalk extensions to that, plus an orthodox file manager and an nvi-clone, and I'll be all set to spend most my time in Opera. :P
OT, I'm not from the UK at all, nor from the US, though I have my fair share of gripes with my gov't.
:P
A power leak is intrinsic to government, or indeed just about any process, I think. As technically savvy people, we know this already. Processing expands to fill available CPU power, memory and disk in a non-linear relation to the amount of work done. Management grows entirely independently of the amount of stuff to be managed.
Also, there is the factor of entropy, which also appears to be intrinsic to higher-level processes, not just lower-level ones. Just compare slashdot now to, say, five years ago.
The problem is not so much getting a large group of people directing their guns at something. It's getting them to agree where the guns need to be pointed, and what needs to be done once the dust settles. You could go out and get yourselves a bunch of gun-toting revolutionaries today, and start killing off government people today, for all I care.
The bottom line, however, remains: there's a *seriously* large support machine at work here. You'll be seen as displacing a legitimate government in an insurgency, so you need to either retain power by force, from which it will be really hard to step down without further bloodshed and possibly retaliation, or you need to turn public opinion, which pretty much entails turning your guns on the media and other support infrastructure. That's some serious carnage, either way.
If you don't commit to that carnage, stuff will bounce back even harder, with people crying out for even more stringent measures than before, to protect their illusion of safety and "normalcy". And if you do, there's at least decades of work to be done in order to stabilize the country again.
Also, there's the whole issue of whether all these gun-toting revolutionaries actually agree on things afterwards. You'll quickly see that you disagree on many points of how things should be run. In the start, when you're working on the simple stuff, everything will go smoothly (for some very stretched definition of smoothly), but once you're getting down to the details of it, there will be a lot of cooks in a kitchen that's getting smaller by the minute. It takes a real strong leader to hold something like that together, and then you're really no better off than before, since it'll be that leader's vision, or it'll all fall apart.
As for the constitution, the real problem is that the population isn't cut out for it. Now, I'm not saying this is a USA-specific problem, far from it. But the constitution postulates an aware populace, which is willing and able to think things through in a Big Picture(tm) way and then follow through on a local level. If you had such a population, things wouldn't be where they are.
Regarding the colony stuff, that's some very different circumstances. Try the civil war or something like that for a better comparison. Or the revolution against the Romanov's. And even in those cases, I suspect the resistance was far less than you'd encounter with a revolution now.
People cherish their lives as they are. They're not all that concerned about liberty or anything as vague as that. They have enough of it that they don't feel it is a pressing issue. Sure, a few will complain here and there, but when it comes down to it, they're more worried about their X-mas bonus and who wins American Idol or whatever.
I mean, come on, how many people here on slashdot have voiced the opinion that someone should take up arms and Change Things Now(tm)? Where's the mailing lists, or what-have-you, organizing this revolution? If you really feel that strongly about it, why aren't you preparing a revolution already? Are you armed? Ready to start shooting fellow citizens? At the very least, you could take a shot at the president or something, send a message that the time to act is now or whatnot.
It's the old frog on the boiler plate again. You're seeing that things aren't what they should be, but that's been the case for a very long
You seem to forget... guns aren't about changing governments, they're about removing them.
Sure, you can remove a government with enough force-of-arms, but how are you going to go about setting up a new government that is better than the old one?
This is a government that represents a significant percentage of the population (26%, I think someone here said). That means 26% of the population is opposed to a revolution, and you'll have to supress them or kill them. Not generally considered a good way of governing a country, now, is it?
This *beautifully perverse* aspect of democracy is too rarely appreciated: in a revolution, there is no longer a single target. A revolution must target the masses. And violence targetted at the masses is something you'll need some pretty hard and pressing reason to make people resolve to do on their own.
Quite simply put: in a democratic society, you can't have revolutions unless *any* alternative is better than the current situation, and for the majority of the population at that.
Social hygiene or potential future catastrophe just won't cut it. It has to be a problem that has *already* materialized, and is significant enough that people are willing to risk their loved ones waging a bloody civil war over it.
It. Won't. Happen. Ever.
Not in the UK, and certainly not in the USA.
There is no established link between Al Quaeda and Iraq. And people pulling out is mostly a matter of not wanting to be in the middle of the cesspool aftermath of the US invasion. You know, civil war and all of that, not to mention the country falling into the hands of religious extremists and being more likely to ally with Iran etc.
And that Northern Ireland is still a part of the UK is a testament to the effectiveness of seeking other solutions than bombing someone you don't like.
Also, as Chomsky has pointed out, terrorism is not effective for the weak. It takes something on the order of an actual nation with the attendant resources to make terrorism work. It's worked great for the USA over the years. Not so for smaller groups.
Again paraphrasing Chomsky, the only thing that was special about 20010911, apart from the media coverage, was that (for once), terrorism worked out well for the weaker party, kind of like David and Goliath. And, of course, that the US response triggered a significant growth in terrorism activity world-wide and gathered more support for the terrorists, although I guess you could file that under "worked out well".
With regards to people taking antidepressants because their lives aren't perfect, I always found that rather funny.
Regular antidepressants (the ones you'll get coverage for in an outpatient setting) don't actually have mood-brightening properties, and many are a bit passivizing. Virtually all of them have some pretty nasty side-effects. So they're perfectly useless for people whose lives just aren't perfect. It makes sense that some people get prescribed antidepressants by ignorant doctors without knowing that they'd be better off with a piece of candy, but I haven't seen any hard numbers on this.
As for major depression, where these drugs have an indirect mood-brightening effect by reducing the depressive symptoms (just as a cure for cancer would have definite mood-brightening effects on the people who have cancer), these doesn't necessarily have anything to do with your life not being perfect. And they do entail your life being, in slashspeak, "teh suck". More so than anyone who hasn't had it, or lived every day for a long time with someone who does, are able to comprehend.
That's kind of been my guide to evaluating whether someone should be on one or not: if the adverse reactions and/or side-effects are something they can't handle, even if the drug has an effect, then they shouldn't be using antidepressants in the first place. It just isn't serious enough to merit treatment, and they need to build coping skills instead, or see a lifestyle guide, or possibly go to therapy (or their priest or whatnot, which does as much good as regular therapy, studies say).
Personally, I've had side-effects that a modern person would consider debilitating by themselves, as well as adverse reactions that were grounds for hospitalization (e.g. circulatory collapse). I still don't see it as a problem: the baseline is worse than even the worst adverse reactions. Compare with cancer patients who get to puke their guts out, lose their hair and suffer lots of pain... if there is any chance of getting better, they generally don't cut the chemo or whatnot.
How about Galactus?
:P
Oh, right, he doesn't live off oil.
Perhaps you should reflect more on that. :)
Killing a superhero with a sniper-rifle sort of makes a statement. It isn't fitting. For anyone.
The realities of our time caught up with him. Cap is dead because the concept for which he was a symbol is dead. And what better way to drive the point home than to be brought low with a sniper rifle?
Yeah. I know that. Which it also states on the website itself.
In fact, a lot of stuff in psychiatry is controversial.
Oddly enough, low-efficacy drugs (SSRIs), zero-efficacy therapy (non-CBT psychology), etc. isn't.
As for the site you quoted, they seem to be missing a few vital points.
For instance, they are criticizing the technology as being experimental. No-one is arguing against that. Call him an "early adopter". Field use in clinical practice is the only way to get research done that is targeted at clinical use, rather than at maximizing profits. Plus, showing distinct differences in organ function can help get people to accept the legitimacy of illness and such. Showing people the reason why compliance is important can also be a good thing.
The language is criticized, blatantly ignoring that the language they are critizising is aimed at patients, not professional. Wordings like "brain balance" are a lot easier for a regular patient to relate to than intricate discussions of the topography of the brain and so forth. Similarly, saying "soul" better describes to the patient what you're referring to than any intricate philosophical treatment of the topic of what comprises the essential features of personality and cognition before launching into the various theories on which features correlate with what in the brain.
I never tell the PHBs what kind of algorithms, schemas, etc. I'm using, unless they ask. Because that's my job, not theirs, and I should explain things to them in terms they can understand, not blather on in techspeak. If they show an awareness of the subject, and an interest in further details, I'd be perfectly happy to level with them, but no way am I going to market myself on technobabble.
For efficacy comparisons, that's pretty simple. If anyone out there that doesn't have a vested interest, but does have access to SPECT, wants to, they should have no problem doing a serious trial. Of course, the fact that e.g. the drug companies don't do this, could indicate that a lot of people are quite uncomfortable with the idea of potentially diminishing their own sphere of influence by indicating that their popular theory isn't all-encompassing.
As for patients doing better, that could be evaluated. The ones I know of have had little or no success elsewhere, but that's just anecdotal again. The identified depression subtypes show a lot better correlation with what I've observed (trait clusters, med response) than any other prevailing theories that I've seen.
False positives and negatives: the critics are probably in a better position to test this. Amen Clinics don't get enough patients fitting regular inclusion/exclusion criteria to evaluate this themselves.
Whether these functional changes are primary or secondary is also most likely irrelevant. Abnormal perfusion and/or metabolism indicates a potential treatment target. Whether the abnormality is due to the illness or causing it doesn't really matter, as long as the treatment works. For research, it matters, of course. A lot. For treatment, the results matter. But, yeah, he should post some more data about any clues he might have about this. My guess is the astroglia are involved in long term downregulation of perfusion and/or metabolism as a result of abnormal activity in those areas, accounting for the slow normalization over time. But some data would always be nice. It would be nice to have the same concerns addressed for MRI, or any of the popular research techniques, too.
But, yeah, David Amen has taken it all a bit far, and has an obvious, financially motivated agenda. So does virtually everyone else. I think the tests are useful in relation to depression. And I have no ties to the clinic, nor have I ever been treated there, nor received a SPECT elsewhere. YMMV.
There is some small core of truth to his statement, although not the way he put it.
/...
Psychotherapy, with the exception of Cognitive Behavioural Therapy, is generally not particularly effective in dealing with "true" depression (which excludes stuff like bereavement, etc), and a "good listener" will do as well as anything but CBT. This has been verified via extensive testing.
MRI is not a very effective tool, compared to PET and SPECT.
Inclusion/exclusion criteria for drug trials leave you with patients you will virtually never see in clinical practice, making the efficacy numbers useless. This also causes issues with women, african-americans, etc.
Drug trials almost exclusively rely on drug-placebo or drug-tricyclic comparisons. It is rare to see a study that is statistically useful, i.e. has drug, placebo, comparable other drug, drug of different class.
SSRIs have no statistically significant difference in efficacy, time to onset, or any other parameter of relevance, aside from CYP enzyme profile. Their efficacy is very limited.
SNRIs have no statistically significant advantages over SSRIs, but more severe withdrawal symptoms, although there is some weak evidence of shorter time to onset.
TCAs are not very viable, due to intolerable side-effects and cardiotoxicity. Their efficacy is variable.
DRIs work very well on several subtypes of depression. Statistically significant advantage in most cases. Fast onset. Abuse potential. Not available, due to the latter point.
MAOIs work very well on most types of depression, if the right type is chosen. One study shows 85% success rate in treating refractory depression with tranylcypromine at doses of 1.3-2.4mg/kg/dy or so. Onset varies from hours to weeks, depending on the type of depression and the choice of MAOIs. However, they require an intelligent doctor and an intelligent patient to be safe, so they are not viable for mass marketing due to the shortage of both.
etc...
I could go on, or even cite this stuff, but this is
Bottom lines:
- Psychologists are often full of crap, and charge exorbitants of money for something they themselves have proven equivalent to a free conversation with your minister or what-have-you. Not counting the CBT people here, of course; there's ample evidence of efficacy for CBT.
- Doctors, and even psychiatrists, are commonly not really up-to-date in their own field, and lack familiarity with any good tools other than what in programming terms would be called shotgun debugging. Doing research in clinical practice is frowned upon, despite being the only way to get real-world applicable results. Using "experimental" tools, such as SPECT, is frowned upon, despite a proven track-record.
- Drug companies develop drugs for profit and damage control, not efficacy. That is, they are designed to do the least amount of short-term harm in the hands of J. Doe for the least amount of production expense. Never mind osteoporosis, asthma issues, weight gain, etc, as those fall outside the scope of the FDA approval process, being long-term and all.
- Simple base research is omitted, particularly where it might threaten the profitability of meds, even competitors' meds.
- Depression sucks, and will continue being the featured syphilitic blow-job of our times for the forseeable future. (WHO projections are rather pessimistic. Mortality rates aren't exactly great, and don't seem to improve. And, like syphilis, a lot can be done to prevent/treat, but isn't.)
So, I guess that's what he was getting at.
Here you go.
http://amenclinics.com/bp/atlas/ch7.php
That's SPECT, not MRI, but it shows this stuff pretty clearly. Compare with some of the other stuff there if you like.
You will not see structural changes. You will, however, see differences in metabolic activity, etc.
And, yes, these can be used as a diagnostic tool, as well as an aid in selecting proper medication to deal with the problem, which will often be necessary by the time you get such a scan.
The condition is not defined by subjective feelings. Have a look at some SPECT scans, for instance.
e -treatment-begins, might have a pretty interesting impact on future developments in the field, particularly with regards to clinical practice and drug efficacy trials.
Subjective feelings are frequently employed in making a diagnosis, however.
There are specific alterations in neurotransmission patterns, glucose metabolism, and other objective, measurable features of central nervous system function. And there are measurable psychomotor and cognitive deficits that develop over time.
Also, the animal models generally reflect a set of fairly objective behavioural features that we can use in working with depression. Essentially, it is most readily induced in animal studies by making a situation hopeless, negative and unavoidable over an extended period of time. You give up, and lie down to die, in a sense. Nothing matters, or is worth doing, because you're just waiting to die anyway, and there's nothing you can do about your situation at your current energy level (or, by the time there is, you have lost too much function to be able to rise to the challenge without extensive rehabilitation).
These are the animal-model features, and their presence can generally be verified in a fairly objective manner. There are also scales that quantify these features in a manner that shows a significantly positive correlation with prognostic markers, etc...
Interestingly enough, many commonly employed scales have very poor severity resolution ability in the to end of the spectrum. When I maxed out a number of more common scales, the MADRS scale had the ability to track variations with a good correlation to the subjective feelings of severity as ranked by people living with me and objective impairment in function and cognition. At all scores from about 20 through 50 on MADRS, I have usually retained a constant near-max score on HAM-D, except my HAM-D improves if I'm too screwed up to get out of bed, or even move, over an extended period of time. Of course, those are the times when being dead would be an objective improvement if only considering oneself.
I think an FDA-mandated scale with a solid mathematical basis (factor analysis, normalization, etc), and good resolution throughout the spectrum from having-a-great-day-every-day to stark-raving-mad-and-needs-weeks-in-the-ICU-befor
There are objective tests. SPECT, for instance.
Have a look at the Amen Clinic pages. They show SPECT images of patients with various mental problems, including depression. Also, they identify several characteristic types of depression, based on the SPECT findings, that respond to different classes of drugs.
The most visible depressions to someone "on the outside", are light depressions (when the mood features are dominant, and there is still energy to bitch about it) and severe depressions with psychotic features (when the whole lost-touch-with-reality, killing-yourself-for-spilling-some-milk, no-rationality-left kind of behaviour takes over).
To me, I've generally been crying more in my better periods, when I have feelings, and actually have any tears left to cry. In a bad period, I can cut myself deeply and not even feel any physical pain, let alone emotional response. If my sister had died during her recent hospital stay, I wouldn't have felt anything. And I've always cared deeply for my family. I would grieve during the subsequent better period, though.
Crying is a good sign that it's not "too severe", unless it is accompanied by psychosis. By which I don't mean to say that spending your days crying in fetal position is a good thing, just that it's about halfway along the spectrum of severity in depression. Which should give people some idea of how bad the really bad cases are, or how comparatively "light" the more "common" depressions are (and how this leads to statistics about full recovery in a few months, etc.)...
Actually, it's simpler than that.
Depression causes a general, progressive decline in cognitive and (later) psychomotor function.
This decline slowly starts to reverse after remission has been achieved.
My therapist did research on this, and the results are pretty conclusive.
Trying to link depression to a hippocampal function phenotype might go places, but I seriously doubt it. And their study appears to essentially not be properly equipped to differentiate general decline in CNS function from anything related to the putative phenotypes. As it is, they're just measuring a decline in hippocampal function.
However, the study can be useful, if taken the right way. Basically, it may allow the development of another objective metric for cognitive decline, which may be useful in diagnosis and determining the severity of the animal-model features of depression (as opposed to the mood related features, etc).
Now, if they could just come up with a better prognostic tool for determining the class of antidepressant that will work for a given patient... Preferrably without needing to resort to a SPECT analysis or somesuch.
Come to think of it, there are some pretty good prognostic markers, but the real problem is that it takes ages before state of the art becomes common knowledge in the research community, and ages more before this becomes common knowledge among psychiatrists, and forever before it becomes common knowledge with the general practicioners.
That, and a lot of patients cannot be treated in the current political environment because they are dopamine-deficient. And the opioid-deficient ones are even worse off, quite often.
Because, that would defeat the purpose of having a handy DDoS-from-hell service around where anyone can anonymously hit tech targets they dislike with 1 day downtime.
This is not a matter of whether they got their monopoly through fair dealing, or even whether the competition is good enough.
It's a matter of making competition *possible*.
As long as they have an effective monopoly, they are not allowed to lock others out of the market. They do not have to actively *help* their competitors, but they may not block them from competing.
Let's take an example. Unless an office suite is virtually 100% compatible with Microsoft Office, it cannot fly, in the sense that it can never achieve significant market penetration. If you build a suite that is 100x as "good" as MS Office, but lacks interoperability with the legacy of the MS monopoly, it will not succeed.
Now, by preventing you from interoperating with MS products, MS is preventing you from competing. Not by doing a better job than you (winning the competition), but by leveraging their position as a monopoly (blocking market entry). The latter is detrimental to the market, and illegal in some areas.
By forcing them to allow competition (by e.g. demanding that interoperability be possible for competitors to achieve via availability of specs), you are not helping the competitors compete, you are making it possible for competition to exist. The potential competitors still have to pull off the competition part of it themselves, building the software, marketing it, etc..
The difference is that with forced interoperability it is *possible* for a competitor to succeed, if they do a good job, whereas without it, it is *impossible* for a competitor to succeed.
Shouldn't be too hard.
Grab the sequenced junk DNA, do some entropy analysis etc to compare it to the functional genome. Anything that turns up highly unusual levels of entropy (indicating compression) would be suspect.
There may be some truth to this, but there is nothing near that level of coordination in the IT world at present. If there were, I'm pretty sure something would be done about it. Fast.
Either way, giving any indication that you are aware of and/or have considered this line of thought, is a pretty surefire way to make them want to tighten your leash, or fire you outright.
How is this in any way awe-inspiring?
It was designed to do this. It was implemented. It worked. Just how any mission-critical software should be.
Any production grade high-availability system will do this. And designing a system to be HA is not as close to brain surgery as a lot of PHBs and codemonkey cowboys make it look.
I quite agree with the parent poster.
The problem is with the "moral core" concept. Basically, humans tend to seperate beings into two groups, with one being "protected" by the moral code, and the other not being protected. Shifting the boundaries between these two groups is an important part of getting soldiers to fight, and the critical component in allowing genocide to happen.
Most people tend to put animals outside their moral core. Familiarity and empathy extends some limited protection and benefit to household pets in most cases, but even these are not usually part of the moral core.
The scary thing isn't that they are doing the same *things* as the Nazis, but rather that they are doing so based on the same *line of thought*.
Off-topic, I wonder how many intelligent sadists fly under the radar by seeking out positions such as these, where they can cook up some flimsy excuse to perform legally sanctioned torture for very little real benefit. I mean, sure, the pigeons can probably be used for something (e.g. delivering a small payload, or performing inconspicous reconaissance), but the technology to do these things, including inconspicous reconaissance, is probably available in China without resorting to this.
... isn't that a corporation ignored potential national security concerns.
What alarms me, is that apparently Lockheed-Martin did.
I mean, if the single SCMM level 5 company could get their shit stolen, then either something is seriously fishy in them not having infosec to match their coding practices, or someone actually told them security isn't an important parameter for this assignment.
What makes you think drug addiction should qualify for this kind of thing?
For one, the dopamine-response associated with CounterStrike is more than an order of magnitude greater than what you get from e.g. heroin.
Second, opening the door to dehumanizing and mistreating *one* group of people means accepting the mental mechanism (moral core) that allows stuff like the holocaust to happen. Now, this isn't meant as a reductio ad hitlerum argument; I'm just pointing out that it's the same mental mechanism that allows the slope to get slippery, and that it's been shown (repeatedly) to be trivial to alter the perceptions that define this border (inside/outside moral core).
Not that I'm doubting that a lot of kids get labelled with ADD, ODD or what-have-you without fitting the criterion for this, but your argument makes no sense.
Giving ADD-medication to a regular kid that's all over the place will make that kid *more* impossible to manage. We're talking about fairly strong CNS stimulants here. They increase psychomotor activity in "healthy" kids. A lot.
Perhaps what you meant to say was that we have an unfortunate tendency to give ADD meds to the kids when they don't actually need it? I can go with that. The problem here is informed consent: they can't give any without having tried it in advance, and that requires having the parents make the informed consent for them, and docs don't always remember to pull the kids back in to ask them if they prefer life *with* the meds or without.
Oh, and, personally, I'd worry more about how Risperdal has been approved to treat irritability, agression and other signs of not being happy, content or "socially well-adjusted" among kids with autism spectrum disorders. That's an antipsychotic, and it might take years off their healthy adult life: statistically 25 years if you take it throughout your entire life. If that effect is one that accumulates linearly, that means the kids lose 1 year of life per 2 years on the drug. I'm not at all convinced this is a tradeoff they'll be happy about later in life.